well, heres my take on the whole situation.
instead of concentrating on porting OS X over to the i386 (a bloted platform anyway) why not port i386 over to OS X. What I mean is, projects like Darwine hold a grand amount of potential in running Windows native apps in OS X. I realize this is a long, long way away, but if apple can maintain both Intel and PPC builds of OS X, I would think that investing a small amount of developers into Darwine (or a similar project) could yeild promising results in a short amount of time.
This may be some of Microsofts reasoning behind purchasing VPC, not soon, but down the road they can market an application which supports running Windows apps right inside OS X without all the need to display the lower levels of windows (it would look similar to what xWindows looks like on OS X right now). I think that is one of the main reasons that Apple doesnt gain market share. I have persuaded many of my computer literate friends to switch, but the ones who havnt, havn't given the standard "Stupid" mac reasons (slow, outdated, etc) but they have given the arguement that "My existing software wont work, I'll need to purchase all new software." Well, that is a valid point, one that could be elimited once projects like Darwine are more mature (I do realize that Wine isnt even very stable at this point, let alone wine on non-intel processors, but if some serious development was put into it, it probably could be)
Also look at businesses. Again, the key is being able to run legacy applications until they decide to upgrade. Sure, the applications might run slightly slower than on an Intel Windows PC. But the Macs are immune from 90% of the viruses out there. The IT department I work in is getting pretty tired of the monthly virus outbreak.
Just think about what I am recommending. The usual arguements that if OS X can run Windows apps, there would be no need to write native OS X apps will probably be brought up. I understand this, but I think of it this way. Apple doesnt make money selling apps. They make money selling hardware. They shouldnt care why you want the hardware, or what apps your going to run on it. They should care about getting you to buy the hardware. If they can get you to buy PPC, and suddenly, PPC is hovering at 20% market share, I think that developers will do some serious reconsidering. Its just that hurdle, those 5 to 10 maybe even years before its reached the 50% mark, then we'd really see some amaizing shifts.