Chip speed: Mhz ratings are only meaningful within a chip family. Comparing between architectures is like comparing lightbulbs based on watts rather than lumens or lux.
ROM protection: Apple was notorious for using custom ROMS to protect its intellectual property back to the days of the Apple ][. They aren't unfamiliar with this practice, but gave up on it because of the performance hit and the ease that one can simply download the ROM to a file and use that. (This is commonly done for emulators like Basilisk and vMac already)
Classic Layer: Rhapsody was the OS without the Carbon layer. Few Mac people wanted it as it was. Few Yellow Box people wanted the blue box. Ripping out Classic would take virtually no time at all. Replacing it's foundation with an emulator might be possible but probably not worth the time and effort of the device compatability headaches.
Red Box: OpenStep was offered in just such a co-existent environment where users could use OpenStep and Windows at the same time. It probably helped sales some, but sales weren't enormous. Was the interest there? That's hard to tell.
Next tried competing in the IA-32 realm as a hardware vendor, as an OS vendor, and as an application foundation provider, but found a hostile marketplace in all of these areas because of Microsoft's dominance. I don't see how bringing Mac OS X back to Intel would be doing anything better than they had already failed at doing.