You can cry about it, you can hold up this or that datasheet from nearly 20 years ago but until you build a time machine that can warp the world back to 1990, that's where any arguments about what CPU is best for the desktop remains.
A fair point, but it seems like for every obsessive fantasizing about what might have been for the 68k or what might still be for the PPC, there's an equal obsessive ranting about how x86 is, was, and always has been The Only Way, instead of just being the one main path (of many possibilities) that the desktop PC market has taken, and from my reading of the OP that's mainly what he's talking about. (Actually, it seems like there's even more of the x86-obsessed killjoys these days than the PPC delusionaries, but that might just be me.)
Yes, it's true that for horsepower x86 kicks any other desktop architecture's ass these days, but that wasn't nearly so true back in the early '90s - the 68k line was still a reasonable competitor, and PPC was a promising up-and-comer. (Though either of these might be more relevant if Commodore had put some real support behind them, instead of half-heartedly sticking in low-end 020s and 030s with a couple MB of fast RAM and letting everybody else provide the workable 040 accelerators.) As someone in another thread pointed out, the x86 didn't really clearly outpace the 68k until the Pentium Pro in 1995.
(IMHO, what the Amiga really needed by the '90s was less a new CPU architecture and more a complete overhaul of everything else. The Z3 bus was a good first step, but the new machines were still tethered to a chipset far slower than even the now-stock CPUs, unless you bought a separate video card and sound card - not good, considering that Commodore was staking basically everything on them.)
