I'm not a science student, and I had no idea what error bars were until I just googled them. I did take AP Chem in High School though, and in college I was a comp sci major before I switched to business. I got through all that without needing error bars, so it seems to me this is something only needed by math or hard science students? What is your actual major? I am actually curious to know who needs this, and I don't mean that sarcastically.
I will try to explain it in simple terms.
When you do any measurement, there will always be errors.
Errors might be systematic (such as your instrument calibration being a bit off), or random (such as you not reading the scale consistently, or current fluctuations in the powergrid causing fluctuations in readout, or your sample being less than representative).
As you can easily understand, these errors become very, very important if you want to compare the results of several measurements. For example, you have obtained a value of 105 under set of conditions A, and 125 under a set of conditions B, for a certain parameter of the system you are studying. Does it mean that going from set of conditions A to a set of conditions B leads to an increase in the value of the parameter you are measuring? Categorically no, because the random error of your measurement is 50 units. For all you know, the value of your parameter under conditions B might be LOWER than under conditions A.
Error bars are a way to convey this idea graphically.
As you can immediately appreciate, error values are as important as values themselves, because they enable comparisons between individual data points, or entirely different data sets.
With no error bars present, such comparisons are completely meaningless.
I'm not disputing that you really need error bars or that they are important to some people, but do you really think they are so crucial to Apple's target demographic that they needed to be in the 1.0 release?
As you might have gathered from my hurried explanation above, error bars are absolutely essential if you plot anything other than, say, your monthly revenue. If you have measured it - there is an error. If you want to compare two measurements and make conclusions, you have to know whether your comparison will mean anything.
As far as target demographic goes...
Macs are used widely by the life science/biochem/chemistry community. Apple does realize the importance of scientists as its loyal customers. For example, they had a booth at every single one of the ACS National meetings I have attended. Yes, right next to Shimadzu, Agilent, and Sigma-Aldrich.
We scientists do a lot of presentations (probably more than your typical businessman), and we love Keynote. At the last ACS meeting I have attended, half of the presenters used Macs. Of those, more than half used Keynote - because it is an awesome presentation program. Better than Powerpoint by leaps and bounds.
If you need more than that, then you should use Office instead and that likely is Microsoft's strategy. iWork isn't meant to be a full replacement for Office, just for some of the functionality for people who would never use a lot of the features.
The only thing is, Office is not Universal - and won't be for 6 more months. I can't do my work with Office 2004 - I get too many spinning beachballs ever since the Intel switch.
There are decent alternatives to Word (Mellel) and Powerpoint (Keynote), but there is nothing that can fully relace Excel as far as I am concerned.
I think $80 is a fair price for what you get in iWork, and I'm happy with the product so far. That's the great thing about Apple offering a 30 day trial- I found out it does what I need, and you found out that it doesn't so you haven't wasted $80 yet.
Since Numbers is useless, and Pages close to useless (due to incompatibility with bibliography software), $80 is expensive for just Keynote. And as far as error bars and resizable markers go - how hard can that be?