The OS would have to be changed not merely so as to run on the Intel CPU but also to make use of the rest of the industry standard PC architecture’s set of chips instead of the Macintosh chips. Otherwise it might “boot in a vacuum” but not recognize video cards, printers, expansion cards, networking devices, etc., and even that is assuming drive controllers and RAM slot architectures were driven and used in a completely identical way, which they are probably not.
That’s still doable. But if they released MacOS X for Intel tomorrow, you could buy it and install it on your PC, but you couldn’t run any apps.
Open-source Unix programs (including X-Window) would be the most readily ported. The real Unix geeks would do it first, then would release source code compilers that would “know” MacOS X Intel as one of the Unix variants they knew how to compile for.
Applications written for the MacOS X “Cocoa” system could be ported by the companies that own them pretty easily too, but they might not bother unless it looked like lots of people were going to buy and install MacOS X for Intel.
Applications written for MacOS X “Carbon” would require a much more complicated recompile; you’d have to go to the source code and rethink the app as a “Cocoa” app. This is also true of “Classic” apps that only run in the Classic environment of MacOS X – no easier porting them to MacOS X under Intel than porting them to Windows. Mostly you would not see any of this happen unless consumer demand made it worth their investment.
Windows applications would not run under MacOS X for Intel except in an emulation environment. You could build such an environment with less of a speed hit than you get with VirtualPC on existing Macs because you would not have to emulate the processor, but you’d still have to run some flavor of Windows as a subenvironment, including waiting for it to boot and so forth, or you’d have to emulate the environment by catching and redirecting all of its calls in a way transparent to the program, which might be faster but probably less compatible.
All in all, such concerns do not bode well for MacOS X on Intel. Lack of software is why there is no BeOS and no XP-on-Alpha as a follow up to NT on Alpha–and lack of software would probably kill MacOS X on Intel as well.
Aside from that, it would not do APPLE any good.