I dont think people realise just how long projects take. When a new product comes out whether its a computer or a shoe, that has been in the design phase for a long long time, a lot of things go into development well in advance.
True enough (and FWIW, I'm also in Engineering).
However, let's not forget that there has been a long history in the PC industry of being agile and quite fast-to-market for things like new CPUs. Similarly, Apple's history with the "Yikes!" PowerMac was a ~6 month development to shelf.
Things we think are brand new and state of the art today have been in development for years, there are things today that were probably being messed around with in research labs over 5 years ago.
True, but his also invokes the question of just how much 'real' innovation is needed here. For example, if the scope of the effort is defined as designing a new motherboard arpund a new CPU and slamming into a cheesegrater case, this is a much simpler & faster { development -and- transition-into-production } cycle than what the Trash Can required.
And overall, the envelope of the next-MP will immediately inform us by how much it diverges from being a brutally straightforward "Truck" for how much Apple chose to push out their own delivery schedule to the right by squandering on superfluous, non-core functions.
Its like the new year model cars, which in reality were in development a few years before they were even in production.
Sure, although it really depends on just how much of a design is changing and by how big of a change that is to one's production lines. Thus, upgrading from a 1TB to a 2TB HDD is s trivial "drop in" that takes but one week after the procurement dept has the new deliveries arriving to transition the line. In contrast, the tcMP threw out the old cheesegrater production line and had to schedule to fab & test out an entirely new manufacturing line.
KISS summary here is that you only take the schedule hit for those things you choose to include. A design which has additional in-house complexity will take longer (naturally), even if it isn't really a necessary (from customer viewpoint) design attribute (yeah, I'm looking at you ... again! Trashcan Mac)
There are things in the world being designed and developed today by various companies that won't see a release date this side of the 2020's.
Sure, but the question here is how much is "development" versus "applied development": the timelines for things like CPUs isn't Apple's but is Intel's; Apple is predominantly buying those components "off the shelf" with the only real schedule constraints being for production quantities at target price points - that doesn't obstruct Foundry samples to fab demonstrators, nor does a PC assembly line have to know what CPU is on the motherboard that gets dropped in; as such there's a lot of parallelism.
As far as the new mac pro goes, Apple stating it is in development in 2017 and saying it will be for next year is not at all an outrageous timeline.
True, it wouldn't be ... **if** it is true that Apple did nothing for 2014-2016 and didn't even kick off the effort until April 2017. However, for this provision, it is appropriate to be outraged that utterly nothing was getting done in 2016 (at the very least), which would have collapsed the timeline down into "overdue today". Overall, any delay that's more than (date that "Selected-CPU" became available in quantity) + (45 days) is an Apple management failure. Even if they went over the waterfalls again with a weird cylinder, etc.
-hh
The sad thing is most of the actual Mac lineup has been stagnant way too long... what have they been doing. I don’t think it’s realistic stick to expect a turnaround in 6 months or even a year or so, but we’re looking at both the mini and the pro, just sitting stagnant-resting on their laurels so to say. Yet development on iPhone X has been 3+ years in the making- - - 
and Windows replaced with macOS, you can kind of understand why they think signing a deal with HP to make it and do just that should be something Apple could do in an afternoon. 