Just came across this thread and couldn't help but to throw my 2 cents in.
Now, this is not exactly what I think *would* have happened, but it's certainly along the lines of what could have happened. But mostly it is what I might liked to have happened.
Atari - I know Jay probably wanted to create his own full-fledged computer, but Atari being savy (sic) knew they needed to stick to there core business and come out with a next generation game machine. Score one for us! A kick-ass game machine that would have kept the Sega competition at bay. As a concession to Jay, and as a way to edge themselves into the computer market allow him a "game-machine" on an ISA card for IBM computers. They would market this not only as a superior graphics card for business and video production, but also as a lower cost way to create games for their next gen game machine. Score two for us! They would have sold like frig'n hot cakes.
CBM - Commodore, not having the whole Amiga thing to derail them, could have focused on their success --low cost home computers. The C128 was an pretty good move, they should have made a little more effort to improve upon that design, but they still had ample opportunity to put out a C256, pushing the hardware to the next level, but with a continued focus on lowering costs. Imagine a C256 that could still run all the old C64 stuff, but with 256K, 8 times faster CPU, and better graphics and only $395 retail, shipped for Christmas 1987/88. At the same time they had slipped a CP/M mode into the C128 that could have morphed into Unix over the next couple of iterations something CBM was already dipping their toes in at the high-end. They should have doubled down on the strategy. It would have payed off well by the 90s.
As you can tell from my analysis and *wishful* thinking. What killed both of them was diverting from their core business and not continuing to build upon their previous successes.
