My recollection is that Android was developed when many were worried that MS was going to find a way to monopolize phones the way they did PC OSes. Lots of people were suspicious of paying a MS-tax to anyone or giving anyone too much control of the platform. I think Google had to give up quite a bit of control to get buzz and rapidly growing market share for the platform.
And that approach has worked remarkably well - http://images.androi...-Market-Share.jpg
With the purchase of Motorola Mobility, the (at least nominal) locking down of the platform - http://www.engadget....-licensees-abide/ - that seems to be changing to some extent (and isn't unexpected at this end).
The iPhone spoiled a lot of people and raised expectations in the smart-phone space. Nobody worried much about upgrading their Razr - they were happy if the flip still worked 2 years later. Now that Apple has shown that you can update the OS on a 2 year old phone (but not without some issues along the way - http://www.eweek.com...-Good-Fit-277996/ ) everyone wants the same thing on their non-Apple phone.
As you say, the smart-phone platform should be updateable for 2 years going forward. But I don't think Google necessarily made a bad choice in not specifying that from the start. The platform needed to be able to evolve quickly. Remember that the original iPhone couldn't run non-Apple apps - http://www.macworld.....cfm?newsid=16926 - things change.
So, it's a tradeoff for users and developers. They want a huge, rapidly growing ecosystem, but they want uniformity. They want the latest and greatest while not being 'forced' to buy new stuff every year or two. Handset makers want a huge market but want freedom to make unique customizations while still controlling costs.
It'll be interesting to see what Google's MM comes up with, and how they'll "supercharge" the platform while still keeping it open.
My $0.02.
Cheers,
Scott.