There are so many other factors as well, like how long it would have taken to develop the AAA chipset to release status. Also, I think Commodore's stance of the Amiga primarily being a toy rather than a producitivity-capable machine would have killed it sooner or later.
Or, x86 architecture and related development would have continued exactly as it has, and x86 would still be the cheaper, better performing option it is now.
And what if the Amiga had clung on to its own hardware technologies for too long, leaving behind the advantages of technologies like AGP has to offer? What is the likeliness of technologies like Zorro slots with only one company's research funding going to be able to match AGP, PCI, PCI-X, etc?
Assuming that Commodore would have made subsequent decisions at the right times, I think they would have ended up with a share like Apple's to the market. Or maybe Microsoft/Apple might have bought them up by then? Assuming that didn't happen, and Commodore/Amiga was still in practical existence and profit today, PPC development may be further along the road than it is now, but otherwise I think the picture would look much the same as it is now.
There are times when new ventures of making new technologies is appopriate, but not all the time. The last decade or so belongs to x86. Occasionally some new incompatible technology will break the mould and legacy of older compatible technologies, but it doesn't work all the time.