Apple Claims 91% of $1,000+ PC Market Revenue in June

[LagunaSol]

But Apple aren't even willing to test that theory. So in the end its just elitism, keeping all the goodies to themselves and charging extra for the privileges.

I'm so tired of this ridiculous "elitism" label thrown at Apple just because they choose not to target the budget-minded consumer. What is this, the Soviet Union? "Computing for the proletariat, comrade!" Good grief.

I suppose Microsoft Office's high price and resistance to accommodating other formats is "elitism?"

And I suppose any retailer other than Wal-Mart is elitist too?

Spare us.

 

[AidenShaw]

Performing overnight backups (as you mention later on) is another utilization example that similarlly makes the 24/7 ROM less unreasonable.

The client wakes from hybrid sleep, incrementally backs up unique modified files, then returns to hybrid sleep. The logs show that this usually takes from 2 to 10 minutes, depending on the system.

The system will often sleep 20 hours or more a day (like my cat). Check mail in the morning, a couple of hours in the evening. Sleeping the rest of the time.

Overnight software udpate checks does too.

Ditto each other 'housekeeping' script that is set to run overnight.

Only if these are set to wake from sleep, and only then if they run for extended periods. (A five second wake to check for updates is hardly measurable.) None of these tasks are set to wake from sleep - the backup is the only WOL-enabled or sleep-timer enabled activity in the house.

The gamer card is a common enough PC expansion and was merely used as an illustration as to how highly variable the power consumption of a generic PC can be: the two cards mentioned consume ~50W at idle all by themselves, so if you were to figuratively add them to anyone's system, your home electric bill will have just shot up by ~$65/year.

Two points:

1. The GTX 285 puts the PC into a different class of system - it's so much more powerful than the Imac in graphics that it's absurd to compare them.
2. Again, only if you disable power-saving and run 24/7.

[*]The reality is that systems (unless they're folding or encoding) spend most of their time in lower power states...

Which can be addressed if you really want to, and was already discusssed. The challenge isn't in making a more 'accurate' model, but if that that model results in a profoundly different conclusion than a KISS ROM.

I disagree with the way that you are making flat statements about TCO, without pointing out that you are loading the model with assumptions that are probably not true for most users.

For example, you said:

Sorry, but the i7 is roughly $1000 more expensive to operate over its lifecycle, so the question becomes one of if the higher performance of the i7 results in more than $1000's worth of productivity gains...."all other factors being equal". For non-demanding tasks ... Internet, Email, etc...the answer is a big fat FAIL: the i7 is overkill for the intended application, so all you're doing is wasting power & money

You declare that it's $1000 more for a Core i7 over an Imac. In my case, the number is less than $40. (The math is (95-50)*4*365/1000*0.15*4). Your estimate is only off by a factor of 25. (It's also a bit confusing that you say that a "generic PC" is $250/year, but you also use $250/year for the Core i7.)

(my 2.93 GHz Core i7 seldom goes above 1 GHz for any length of time during normal use).

That you bought too much machine for what you actually use it for is an interesting tangential point, since it counters the common argument that the iMac is underpowered and suggests that such "overbuying" may be a common behavior within the PC community.

There are times when I do video encoding, or file compression, or run multiple active virtual machines, or... and the Core i7 is much "snappier" than a laptop CPU.

I'm willing to possibly "waste" $10/year on electricity to save hours of processing time on occasion.

There's an obvious automotive analogy - did I "waste" money on the turbocharger on my car, just because I only open the wastegate a couple of times a week? The answer is "no".

Your Q6600 system with 6 disk drives serves as a poster child that illustrates my point.

What point does a system with 1 spinning drive and 5 drives that power management has spun down make?
__________________________________

My main point is that systems have become much more aggressive in power management, and default settings from the manufacturers are also greener.

I feel that your "approximation" of taking the worst case TDP of all components, and assuming full load 365 days a year, is not the out-of-the-box experience that people are getting.

Systems do not run at TDP except under extreme conditions, and the rest of the time power management is slowing clocks and shutting down cores even while the system seems to be fully active. The hybrid sleep state has replaced the screensaver as the "idle" state.

Some of the peer-to-peer technologies are even struggling with this "problem". As new systems replace older ones, and are set to spend more of their time sleeping, the P2P applications are having to provide more redundancy or new features in order to function. (See A Prototype Power Management Proxy for Gnutella Peer-to-Peer File Sharing (pdf), for an example.)

In short, I don't buy your argument that an Imac would save me $1000 in electric bills, period.

 

[twoodcc]

wow. i guess the small price drop really helped. i hope they do the same to the desktops soon

 

[ilovefanboyz]

I smell a lot of jealous in this thread.

You do realise that there are about 30 PC's sold for every 1 Mac?


What is there to be jealous about?

Again, somebody who thinks this "90% market statistic" is meaningful needs to get their head out of the sand. Re-read the conditions of the statistic.

 

[Carniphage]

You do realise that there are about 30 PC's sold for every 1 Mac?

What is there to be jealous about?

Again, somebody who thinks this "90% market statistic" is meaningful needs to get their head out of the sand. Re-read the conditions of the statistic.

Look again at some numbers.
By unit share - Apple sells less computers than three companies.
Dell, HP's consumer division and Acer.

Apple makes more profit than all three *combined*.
Now Apple is not just a computer manufacturer. - So let's try to remove Apple's iPod, iPhone and software profits by halving Apples overall profit share.

Apple *still* makes more profit than all three combined. That statistic reveals just how broken the PC hardware business is.

There are a lot more Windows computers out there than Macs. But that's part of the problem.

The competition in the PC space has effectively robbed manufacturers of the ability to make a profit. This complete lack of profits has taken away the manufacturers willingness to innovate. They don't even try to add value. Instead they look to make micro savings, making cases thinner - buying cheaper connectors.

For some consumers low prices are great news. But some consumers look for value, not just cheapness. And it is those consumers who are turning to Apple in increasing numbers.

C.

 

[-hh]

You're going to need to change your baseline for a gamer card. The GTX285 isn't a gold standard that you can slap down and say everyone and their dog has one...

I understand your point, but my point in mentioning the PC gamer cards was that regardless of which particular model they are, it illustrates that PCs are more prone to being "nickel-and-dimed" in power consumption.

Its easy to pin down how much an iMac uses .. there's an Apple webpage for that. The challenge is in trying to figure out a fair 'average' to assume for a million invariably customized desktop PCs. That's why the original quick ROM was merely assuming a 200W differential, based on respective PSU sizes.

You're going to need to hit the 8800GT G92 and its derivatives as being a proliferate base or a Radeon 4850. Even then the 2D idle or 3D performance states are set by the vendor's BIOS and can be modified by the user. My 4830 idles at 160/250 MHz which is lower than the base 500/700 MHz of other cards.

Sure...just provide a reference that shows a test of their actual power consumption. The only reason why I used the cards that I did (and there were others also listed) was that it actually had power consumption data. While I only mentioned 2 video cards by name, you can go look at the results of all of them and you'll see that I was picking a reasonably fair representation with the aforementioned rounded-off values of 50w & 150w.

You're trying to exaggerate the power consumption. At least it wasn't something as far out as a GTX295 or 4870X2.

No, I'm just trying to assess if an average 200W difference was reasonable. If its not, then pick another number. For example, even only a 100W difference works out to $130/year and thus is $650 - $910 across a 5 and 7 year lifecycle.

I'm not trying to "Rig the Playing Field". Just merely trying to see what is a reasonably fair number to use based on facts without getting mired into a 25 page analysis ... its just a ROM, with all of the inherent limitations of it being a ROM.

I'm sure most iMacs are in their idle states just as much as someone sporting a Core i7.

Yet Aiden's power consumption number says otherwise (~110w), which was provided before Apple's published value (~60w).

...and if we go with these, as merely a 50W difference and assume the extreme low end case where both computers are merely plugged in and left to idle 24/7 for 5 and 7 years, then the power consumption differences are $330 and $450.


-hh

 

[AidenShaw]

Yet Aiden's power consumption number says otherwise (~110w), which was provided before Apple's published value (~60w).

You love to round up the numbers for PCs, don't you? Why pick the upper end for my Q6600 instead of the 95watts for the Studio XPS Core i7?

...where both computers are merely plugged in and left to idle 24/7...

You mean:

...where both computers are merely plugged in, power management disabled, and left to idle 24/7...​

Because if you don't change the power management settings, the Vista PC will only use a total of $26/$37 worth of power in 5/7 years.

And if you adjust the power management to shift from hybrid sleep to hibernate after 24 hours, the Vista system will cost $3.30/$4.66 for 5/7 years.

But, to me this exercise points out how unreasonable your 24/7 claim is - when the system is able to do much better if you leave it alone. If you're really concerned about electricity costs, you won't be leaving it on 24/7.

(By the way, our corporate PCs have power management policy set by domain policy, and it's

• 30 minutes - monitor off
• 2 hours - sleep (two hours, so it won't go to sleep when you're at lunch or a meeting)
• 8 hours - hibernate (so it won't hibernate during the day)

for about 20,000 desktops.)

 

[BaldiMac]

yeah, i am sure PC's have 90% market share envy... At the end of the day, windows dominates the business space and until apple even scratches that, this just speaks to their refusal to move off their absurd margins.

At what point do you consider a margin absurd? What's the number?

 

[AidenShaw]

Dell Energy Savings Calculator

Dell Energy Savings Calculator (web page calculator)

This page lets you estimate actual yearly power consumption for systems, based on measured power consumption and user profile. (The profile considers lunch and breaks, office productivity apps, and high performance application usage for certain intervals during the day.)

It comes up with $36.77 per year for an Optiplex desktop with a 20" LCD panel. I don't see how you can save $250/year when you're only spending $37 .

This assumes 7 hours/day for productivity apps, 1 hour/day for power user apps, one hour lunch, and two 20 minute breaks. You can adjust the profile settings, and select different client systems.

Even when I set it to 24/7 for performance apps, it's still only $137/year total - or about half of -hh's estimate of "savings*.

Read the Dell Client Energy Calculator Methodology Paper (pdf) for an explanation of the scientific method used to estimate the power consumption. It's far more sophisticated than anthing we've discussed here. (It seems to focus on XP and Core 2 systems, so improvements in Nehalem and Vista aren't considered.)

 

[Goona]

You do realise that there are about 30 PC's sold for every 1 Mac?


What is there to be jealous about?

Again, somebody who thinks this "90% market statistic" is meaningful needs to get their head out of the sand. Re-read the conditions of the statistic.

Right but not all those PC's comes from one company, it is spread out amongsgt many different companies so of course they will outsell Apple but Apple is still smiling all the way to the bank. Acer for exampel makes profits of 60 million on revenue of 3 billion, pathetic.

 

[drsmithy]

Check Made out to Apple, Inc. still uncashed...

While I agree this is good news for Apple's profit margins, it's bad for some people wanting to buy a Mac, especially in a bad economy and especially for some potential PC switchers.

I'm still waiting for that somewhere around $1500 range mythical mid-range Mac tower.
A check is written and waiting to be cashed, only the product does not exist!

I suggest you invest that money. It'll be a cold day in hell before Apple releases a machine that will trash their Mac Pro sales.

Heck, buy some Apple stock with it. At least then they're likely to give you _something_.

You still don't see that many 3.0Ghz or faster processors today; builders are using multi-core technologies to get higher 'equivalent' speeds. Core2Duo, and now tri-core, quad-core and even 8-core and 16-core machines. They still all run at or below 3.0Ghz and bus speed has become the limiter.

Those processors are much, much faster at the same clock speed. A Nehalem CPU is 50-100% faster at the same clock speed than a Core 2 CPU.

So I'd say Macs are high-end equipment by comparison to the sub-$1000 PCs.

If you're talking about performance, most Macs are decidedly low-end. You can buy sub-$1000 PCs with ~4x the performance of the cheapest ($1200) iMac. A PC with equivalent performance costs about half as much. The same is true of the Mac Mini and even, to a degree, the Mac Pro.

That same iMac happens to also be a more energy-efficient system because of that "lower performance" portable CPU: at $0.15/kWh, it typically costs roughly $250/year less electricity to operate than the generic desktop PC, which means if that 'better' PC cost $1000, the iMac will have a time-until payback (comparative break-even) at a mere 2 years, whereupon it then becomes the less expensive system.

Where are your power usage figures coming from ?

I'm going to take you at your literal word here. You've never run a malware scan. You've never run a virus scan. You've never had to defrag your drive. You've never had to run a chkdsk. You've never had to clean out your registry. You've never had to do anything to maintain your computer.

I havae certainly never had to do any of those things. I do, on rare occasions (maybe once or twice a year) run one of those online virus/malware scans on my machine "just in case", but in all the years of doing so, they've never picked anything up.

 

[-hh]

Where are your power usage figures coming from?

Its merely a delta of 200 watts, * ($.15kWh) * 24* 365 = $262.80 /year.

For the iMac, the PSUs are 180W and 250W for a 20" and 24"

For the plain vanilla desktop PC, its hard to gage what is "average" because there's so much variation. Dell's XPS i7 is 360w (before monitor), which would then be easily ~400W, and I have coworkers with systems that are at 600W and 850W. Since this was only a ROM, I went with a 400W baseline, so 400W-200W = 200W differential.

Comparing PSUs have two basic shortcomings. The first is if PSUs are unrealistically oversized, resulting in a misleading value. Due to the PC market's extreme competition and narrow profit margins, this is unlikely. Apple is the same conclusion for slightly different reasons ("slim", etc).

The second question is then what percentage of the rating is the average expected power draw. There is no utterly correct answer here that's still a simple ROM and generalized case. To run at the full rating is (as has been said repeatedly before) a high side worst case scenario, but there's also multiple sources of under-estimation that compensate ... as has also been said repeatedly before.

The KISS bottom line is that there is most undoubtedly a power consumption difference between the "laptop components" iMac and a traditional PC desktop. The only question is magnitude. Because of the great degree of variability in PC configurations, as well as personal utilization, this will invariably be a personal YMMV.

Just be wary of some of these personal YMMV-based claims. For example, one can gage the credibility of a poster who claims that their system burns no power because it is at idle for 20+ hours/day ... even though the timestamps on their many, many MacRumors posts clearly suggest a different reality.


-hh

 

[kate-willbury]

Yeah, but every Apple Laptop but one is $1000+, so is capturing 91% of that share really a surprise? It's basically their entire lineup competing.

seriously. there are hardly that many laptops that go beyond the $1000 price point.

it might as well say "apple stores sell the most apple products! WOW!"

Its merely a delta of 200 watts, * ($.15kWh) * 24* 365 = $262.80 /year.

For the iMac, the PSUs are 180W and 250W for a 20" and 24"

For the plain vanilla desktop PC, its hard to gage what is "average" because there's so much variation. Dell's XPS i7 is 360w (before monitor), which would then be easily ~400W, and I have coworkers with systems that are at 600W and 850W. Since this was only a ROM, I went with a 400W baseline, so 400W-200W = 200W differential.

Comparing PSUs have two basic shortcomings. The first is if PSUs are unrealistically oversized, resulting in a misleading value. Due to the PC market's extreme competition and narrow profit margins, this is unlikely. Apple is the same conclusion for slightly different reasons ("slim", etc).

The second question is then what percentage of the rating is the average expected power draw. There is no utterly correct answer here that's still a simple ROM and generalized case. To run at the full rating is (as has been said repeatedly before) a high side worst case scenario, but there's also multiple sources of under-estimation that compensate ... as has also been said repeatedly before.

The KISS bottom line is that there is most undoubtedly a power consumption difference between the "laptop components" iMac and a traditional PC desktop. The only question is magnitude. Because of the great degree of variability in PC configurations, as well as personal utilization, this will invariably be a personal YMMV.

Just be wary of some of these personal YMMV-based claims. For example, one can gage the credibility of a poster who claims that their system burns no power because it is at idle for 20+ hours/day ... even though the timestamps on their many, many MacRumors posts clearly suggest a different reality.


-hh

sorry but this is silly. my desktop pc's power supply runs at 220w. it obviously completely depends on what the user buys. if they were that considerate of how much power their pc will take, then they will take the care to purchase the right computer/parts to build one for their energy needs. saying apple computes run cheaper than pc ones is ridiculous.

 

[Eidorian]

Just be wary of some of these personal YMMV-based claims. For example, one can gage the credibility of a poster who claims that their system burns no power because it is at idle for 20+ hours/day ... even though the timestamps on their many, many MacRumors posts clearly suggest a different reality.


-hh

My processor sticks at 800 MHz and my GPU at 160/250 MHz when posting on MacRumors.

Web browsing isn't going to cause my hardware to leap into full performance unless it's flash and even then I can force the idle power state regardless of load. I'm not at home so my tower is asleep at least 23 hours a day.

 

[kate-willbury]

Huh? I mean there are plenty of >$1000 offerings from Alienware, Toshiba, IBM, Dell, etc. It isn't like they're competing only against themselves.

This IS a pretty big deal - and points back to: people who know what they're getting and care about what they're getting buy a mac. Those that can live with something cheapo because they don't know the difference, or simply don't care about the difference don't buy a mac.

And by the way, that isn't meant to be denigrating at all to those users who don't want to pay the price premium. I don't care a lot about high-end performance in my car, so I own a Saturn and a Chevy. Both of which are cars that get me from A to B reliably. People who know and care about performance are probably buying Porsche, Jaguar, Mazerati, or something else. That's fine, if they point out that their car is better designed and better built than mine, I'll just agree and move on. I haven't been insulted, they simply pointed out a fact, that probably didn't matter to me anyway.

Same with computers - high-end $2000 computers are going to be better than a $350 computer. Maybe in ways that you don't care, but they are better. What's interesting here is that once you decide you want and care about a >$1000 computer you are in a staggering minority if you buy anything other than a mac.

nice try. just because a mac is more expensive does not mean its the 'ferrari' of the computer world. i can run mac osx on a pc with the right parts just fine. i've had hackintosh 10.5.7 running on my amd triple core, 4 gig ram machine for the past 6 months. that computer cost me $420. inherently, the parts inside are basically the same as any mac. while the interior of a ferrari vs a chevy is obviously VASTLY different. this is the problem with certain mac users imagining delusions of grandeur in their heads.

i love my macbook, but i don't get any 'higher performance' from it then i do from my much lower priced laptop. its silly to think just because we decided to overpay for something means we have some bratty privilege to spew nonsense about the superiority of macs.

i would never buy a pc over $1000 because i can easily get one for my needs for MUCH lower.

 

[Shasterball]

I guess if you are going to spend good $$ on a computer, it might as well be an Apple!

 

[revbarabbas]

See this post is the only reason that I have even tried to stir up a fight. Fine hate me for what I say but please think of the children!! This kid thinks apple uses some fancy magic computer parts in their systems that are better than all other computers. Other than the case THEY ARE THE SAME PARTS!!! Most of the time they are inferior to the $800 computer.

First of all I am not a kid. Second Apple has consistently used the highest quality parts over the lowest bidder unlike the PC market which will use the lowest bidder over quality. It's a matter of company policy.

And how do you know that the parts are inferior? Now I'm no idiot...I use PCs and Macs, I program for the PC and the Mac, and I build PCs so I know the price of a Mac is comparable to the value of the parts that go into them...a PC that is setup like a Mac Pro would cost just about the same except that with the Mac you are getting more high end parts...not all electronics are made the same...a 10 cent resistor is going to be cheaper than a 50 cent resistor but will burn out twice as fast. Plus design is a part of the price of a Mac, an OS built for the hardware is better than an OS built to cover a wide range of generalized hardware, and I'm not Windows bashing...I'm also talking about Linux and FreeBSD et. al. People who know Linux will tell you that ALSA is a nightmare with most soundcards, and the video drivers are a nightmare also.

So to rebutt your statement "Think of the Kids" I am and I would rather have a quality built Mac than a cheaply built PC for my kids, which is why they all have Macs, I have baught 7 PC laptops in my day, they have all had hardware issues under 5 years, but my 10 year old Powerbook still works, battery still has a charge and it still gets used to develop for OS 9. Hell I still have an Apple IIe that still runs, my TRS 80 and my old 286's died years ago. So Apple is simply built better and I know this from experience as well as simply listening to the common sense adage "you get what you pay for"

The difference is that PC users can pick and chose what parts go into a computer. I for one build my own. When is the last time you built a Mac? Can you chose what brand of power supply, cooling fans, processor, what brand of graphics card manufacturer, etc. Goes into a Mac? There is a whole aisle of sound card vendors at Microcenter that one can purchase and pop into a PC - along with a million other upgrades. THAT is the difference.

So you've built a laptop then? Because thats what we were talking about. Maybe you should have checked the frame of reference before you make an obviously idiotic post. And I have built many PCs...the hardware is much broader for the PC market...thats why it's the PC market but for the Mac desktop there are also a lot of upgradeable parts...like the fans, videocards, and soundcards...there just aren't as many vendors for them because there is only one Mac Tower Desktop.

(by the way the case is the only part that apple uses superior material on)

Not true...simply not true and by you saying that is admiting your own ignorance...I have taken apple laptops apart and pc laptops apart and the quality of the simplest parts such as capacitors is usually higher quality on a mac as opposed to the pc.

 

[Eidorian]

First of all I am not a kid. Second Apple has consistently used the highest quality parts over the lowest bidder unlike the PC market which will use the lowest bidder over quality. It's a matter of company policy.

Apple is using the same OEM components like everyone else. ASUS or Foxconn make the logic boards.

 

[revbarabbas]

*BZZZZZZZZZZZZZZZZZZZZZZ* Wrong answer!

Apple does something the other brands don't do, except for the high-end builders like Alienware; they actually specify the parameters of the parts they order and if those parts don't meet the specification, the entire shipment is returned to the supplier. How do I know? I used to work for one of those suppliers and I've seen a shipment returned because a random sampling of components fell outside of Apple's specification; forcing a 100% QA retesting before Apple would accept another shipment.

Maybe you should check out iSuppli. Their teardowns of computer products have reported more than once that "...with Apple you get more bang for your buck."

You said it brother!

 

[shawnce]

Apple is using the same OEM components like everyone else. ASUS or Foxconn make the logic boards.

Well in many situations that is true however...

Apple often uses better case materials, internals and superior machining. Apple uses custom battery technology (in current generate laptops) the goes well beyond what other vendors use in terms of density, form customization factor and charge cycles. Apple has set higher standards for LCD panels, in most situations, then other vendors.

In general Apple sets higher requirements on the component they purchase.

 

[Eidorian]

Well in many situations that is true however...

Apple often uses better case materials, internals and superior machining. Apple uses custom battery technology (in current generate laptops) the goes well beyond what other vendors use in terms of density, form customization factor and charge cycles. Apple has set higher standards for LCD panels, in most situations, then other vendors. etc.

I don't have any complaints on the cases or batteries (that depends on their vendor too).

The LCDs have been terrible until recently on most of their notebooks.

Seriously though open up an iMac and tell me all the caps are solid. Everyone was hit with bad caps back in 2004/2005.

 

[revbarabbas]

Well:

1)Macs are high end compared to the average PC. This is not true. Apple fans think this is true but they are wrong.
2)The 3.0Ghz barrier. Is not true. For mac it might be.
3)Mac uses better parts than PCs less than $1,000.

And just what are your sources for this info? Because you never seem to cite any sources.

I differ with your beliefs but could you point me to the link that shows apple's special high end parts.

There are no links that show the layout of the entire logic board but I can tell you from personal experience in taking apart a Macbook Pro that the individual chips and capacitors and resistors were of the highest quality. Excellent specs on all my electronic diagnostic equipment.

 

[-hh]

Web browsing isn't going to cause my hardware to leap into full performance unless it's flash ...

Same here. However, the ability of modern PCs to multitask means that people will invariably use that feature. There's already been some contemporary examples of multitasking mentioned, such as grid computing, automated data backups, software updates, checking email, etc...even though they may appear to be small individually, they nevertheless still add up. Add to that DVD ripping, video encoding, rendering, compilers, macros, etc...the more that you've found that you can use your machine for, the longer the list of 'background tasks' tends to get.


-hh

 

[revbarabbas]

Apple is using the same OEM components like everyone else. ASUS or Foxconn make the logic boards.

And they use the highest quality logicboards from ASUS and Foxconn...made to their own specs. Instead of the cheapest ones.

 

[drsmithy]

Its merely a delta of 200 watts, * ($.15kWh) * 24* 365 = $262.80 /year.

This is a mind-bogglingly wrong way to compare. It's like comparing the fuel consumption of two vehicles by looking at the highest number on their speedometers.

In fact, it's so wrong, that the only logical conclusion is that you either lack the understanding to be making such a comparison in the first place, or you're trolling. Your "estimation" that the average PC has a power-sucking video card rather than an integrated one, further reinforces this point.

For the plain vanilla desktop PC, its hard to gage what is "average" because there's so much variation. Dell's XPS i7 is 360w (before monitor), which would then be easily ~400W, and I have coworkers with systems that are at 600W and 850W. Since this was only a ROM, I went with a 400W baseline, so 400W-200W = 200W differential.

A fair way of gauging the average PC's power supply - which you obviously aren't interested in - would be to have a look at a few PCs that fit into the ~$600 price range most people buy at. If you did that, you'd find that 250W is a more reasonable assumption.

Just be wary of some of these personal YMMV-based claims. For example, one can gage the credibility of a poster who claims that their system burns no power because it is at idle for 20+ hours/day ... even though the timestamps on their many, many MacRumors posts clearly suggest a different reality.

I don't need to be "wary", because not only have I been in the industry long enough to know about how much power the average modern PC burns from taking actual measurements, but I also have a sufficient understanding of PC internals to be able to make accurate and fair estimates in the absence of hard data. I know when someone is making reasonable guesstimates and when someone is spewing ********, and your posts are the ones reeking to high heaven.

Your assumptions and calculations regarding power use are both wrong, and dishonest. It is unlikely that the average modern PC pulls over 100W, even in "peak times" and is likely around 75W during "typical usage". If I had to make a highball estimate for the average power consumption of a modern PC over it's lifetime, I'd say around 90W. The idea that an iMac (or Mac Mini) is going to save meaningful amounts of money in electricity over a typical PC in its lifetime is laughable at best, outright deception at worst.