i tihnk thier is a good chance this well happen beacuse this will get all the peecee userrs who want to switch over to the mac stop questioning compatibility. I used to question that as well, but now I'm happuily on a mac, and I did this way beofre lindows. Not only will we get M$ compatibily but also Linux. This would be a great investment for Apple if they make it, but you cant forget the lawsuit on Lindows from M$.
I don't see the software coming to the Mac... especially when there is VPC out there. I can see VPC coming packaged with lindows, if it ever gets released, and works well.
They would have to rewrite the code to get it to run on the Mac hardware. Granted, that would have less variables, but would they do it for less then 5% of the market?? M$ has embedded itself so deep into the peecee market that it would almost take an act of god to get it out of there. It doesn't matter if something else is more powerful, easier to use, or cheaper. M$ will either bury the company, or suit it to death.
Technologies such as Wine and Lindows are reliant on the x86 architecture to make the Windows environment to run fast. Its the same for MOL (Mac on Linux) and Classic environment; they won't run on x86 architecture, only Mac PPC.
Think of these things as facilitators, not virtual implementations of architecture. The way Virtual PC works is by writing a whole x86 machine in code. This way a virtual x86 processor runs on Mac OS and a PPC architecture. This requires faking out all Windows programs you want to run by making them think the Virtual PC program is actually hardware.
MOL, Classic, Wine, and Lindows operate as facilitators. Instead of rewriting the hardware in software, these programs simply channel hardware instructions to the actual hardware being run on. So when a Classic program executes an instruction, Classic simply passes that instruction along to the real processor, gets the result and passes it along to the Classic program. So facilitators still fake programs out, but instead of handling the instructions in a virtual processor, the passthe instructions along to a real processor.
There are huge performance benifits to running as a facilitator. Hardware is *always* faster than software. But to port Lindows or Wine to the Mac would be impossible: Macs don't run on x86 hardware. That means that every instruction would have to be translated from x86 language to PPC language on the fly. What would that be? An emulator...a virtual processor. And not only would this port take you a long time, but you'd be back to where you started: a virtual processor. This would have absolutely no theoretical performance gain over Virtual PC.
So, in conclusion, it will never happen. If someone wanted to make Windows programs run on Macs, they would approach it the way Virtual PC did. The alternative is just too difficult and without benifits.