I'll be honest, I think this move is a potential disaster for Apple. The pitfalls are many, the upside mostly mundane and, where not mundane, hard to see clearly. It's scary, but the most promising analysis I've read (Cringely) has Intel buying Apple to give away the OS and kill Windows! Perhaps the best approach is just to look at the plusses and minuses on each of the many facets of this situation.
Firstly, there's the much-quoted Osborne effect, where Apple sales dry up because nobody wants to buy the now obsolete PowerPC-based machines, while the new Intel-based machines are not yet ready. Time, and the next few quarterly sales figures, will tell, but I would be surprised if there wasn't some truth in this. That said, this is the least interesting facet, so let's leave it at that.
Secondly, Apple are now at the mercy of IBM for the next two years. How far do their contracts extend into the future? Are they expecting that IBM will be more, rather than less, likely to mess around with their production lines and their scheduling in order to please a client that, from now on, just doesn't matter? If they haven't got the most watertight of contracts covering the next 24, better make that 36, months, then their mid-term future likely contains chip shortages (assuming that they are actually shipping any boxes -- see above). IBM also has the power, pun very much intended, to make Apple look foolish indeed if, by the time Apple make the switch with their desktop boxes, cooler, faster POWER chips replacing the 970 have appeared that, just coincidentally, beat the pants off Apple's new Intel offering.
Thirdly, Microsoft. One half of Wintel. The major half. Only now it is one leg of a tripod. The Eye of Bill must be turning towards Cupertino, rimmed with fire. In the past Microsoft repeatedly threatened Intel when Intel tried to expand into areas that Redmond felt were theirs by adding firmware to the CPUs, and Intel repeatedly backed down. Then Microsoft flexed their muscles and showed they had choices of their own by porting Windows -- to MIPS, and to POWER, though one was never quite sure how serious they were about actually selling licenses. Once Intel came back into the fold these ports were allowed to die a natural death. More recently though, they've seen Intel investing in Linux and not been able to smack them down because, one supposes, of anti-trust concerns. So the porting gambit comes out again: Microsoft now has an OS that runs on triple PowerPCs; sure it's mostly optimised for games, but those machines are also going to be networking, instant messaging, authenticating identities, and doing much of what ordinary, everyday PCs do.
It has already been noticed that, with the new XBox, Microsoft are throwing down the gauntlet to the low end PC market, saying in effect: we now have our own low-end PC, it's a different architecture from anything you're used to, and it's tightly controlled by us and you're just not going to get a look in; say good-bye to your consumer sales, they are ours now. The point is of course, that that gauntlet also hit Intel on the way down. Indeed, it's possible to see Intel's flirtation with Apple as a response to the new XBox: if you don't need us, well, we don't need you either. In which case, Apple has wandered in just as things are about to get messy. Bill is quite capable of going to Intel and demanding that Apple be taught that it comes second to Microsoft in Intel's affections. Intel may or may not give in. If it does, Intel gets a few more months or years of peace while Microsoft prepares to jettison them forever, and Apple gets screwed. If not, Apple suddenly gets much more important to Intel, and Cringely's prediction may indeed come true. In this bizarro world, Microsoft and Apple swap hardware partners and the rest of us are left wondering what's going on. The final outcome would probably see Apple exiting the PC business, Intel giving OS X away with its chips—maybe even donating it to the FSF—and Microsoft heavily beholden to IBM. Ho-ho-ho.
Fourthly, comparability. Here there are some truly serious risks for Apple. It's not so far been possible to strictly compare OS X against Windows because of the fuzziness emanating from the hardware differences. Apple has always been able to claim that Altivec blah, blah... and so on. Now however, magazines and testing labs are going to be able to build an exact Windows counterpart for any Apple machine, differing only in OS and BIOS and a few, unimportant bits of BIOS-related hardware. For the first time it will be possible to put OS X up against Windows on essentially the same hardware, and compare the two OSes directly. Apple had better be sure that OS X compares favourably when this happens. Otherwise a significant proportion of their desktop hardware sales will have the OS ripped out and replaced by Windows, and Apple turns into essentially a niche Windows system builder offering nicer boxes. Development for OS X will collapse and, again, it's just a matter of time before Apple exits the business. In this respect, it's interesting that OS X rests on a version of BSD that, reportedly (i.e. I've read a couple of articles saying so, but I'm no expert myself) may have problems with threading performance, certainly compared to Linux, probably compared to Windows.
The most interesting thing about these scenarios is, in none of them do things continue as they have so far, with Apple owning a small and slowly declining share of a rapidly expanding market. Things have become discontinuous. The possible futures are wildly divergent (though it's interesting to see that "Apple exits the PC business" is the outcome of more than one of them), and seem mostly to revolve around Apple going for broke to get a much larger share of the market. One thing that can be stated with certainty however, is that though the recent announcements have tried to de-emphasise it, Apple is now pursuing a much, much higher-risk strategy than before.