A few points:
1) Apple are seriously cashed up. And I mean *seriously* -- thanks to past efforts and the runaway success of the iPod, Apple could continue operating for ten years without selling another Mac. There were hiccups in the 68k-PowerPC transition, there will be bumps here, too.
2) Apple learned from the mess that was Motorola, so they weren't game to totally trust IBM, so project Marklar -- OSX on x86 -- was the 'hedge bet'. The decision to make Darwin LGPL open source was part of this.
3) Even though IBM might not have been able to keep Steve happy with technology advances, they've been able to woo all three major gaming console makers to PowerPC. This means that PowerPC is not going to die any time soon, and as such we will continue to see PowerPC-based Macs on Apple's product list for several years yet. At no point did Steve say Apple are switching *away* from PowerPC.
4) There are many many businesses out there relying on Macintosh as their primary computing tool, and who upgrade their hardware annually to ensure the fastest, most productive work environments possible. This 'switch' will not affect their choice: they will continue to buy the fastest-possible Macintosh, no matter what it is.
5) The companies that *should* be worried are those that make name-brand commodity PCs for general use. With the potential lure of having a Macintosh that can dual-boot both major OSes only a year away, especially one that would cost a 'mere' two or three hundred more than a comparable Windows-oriented machine, people will hold off.
I personally don't see troubles ahead for Apple. They've done this transition thing before, and they've got an incredibly canny CEO who is capable of seeing the bigger picture and making apparently risky moves without seriously endangering his company.
Good times ahead :)
I still think that the move to Intel is a Good Thing, and more than just a hedge-bet move ... whilst at a low architectural level ia32 is demonstrably inferior to PPC, for most people this point is totally irrelevant. What it does is present choice against the ia32/MS hegemony, something they didn't have before.
We know that Apple will rigorously defend their decision to keep OSX running solely on Apple hardware, and also that unless Apple do something extremely radical architecturally, there *will* be 'hacks' to allow OSX on commodity PCs. But I also predict that Apple will eventually (and sooner rather than later) help MS produce 'VirtualMacintosh' -- copies of Office Pro for Windows will come bundled with a licenced copy of OSX.
The 10.4.1-for-Intel "leak" (which turned out to be nothing but a practical joke -- run it and you're presented with a gaping anus) could have become, as one site said, an incredible piece of viral marketing for Apple. I'm certain that SteveCo is avidly following the developments of this, and the feedback & responses that it is generating, so it would not suprise me if SteveCo eventually decide to do it themselves. As Leopard-for-Intel approaches completion, 'Tigertel' will get a little tidy-up, have iLife removed, and a few other things trimmed or curtailed, and it will be released as "OSX Trial Version".
These issues aside, there will still be one very large problem to overcome: program operability. Even though you can run a new OS on the same box you used to run Windows on, you won't be able to run programs developed for Windows under OSX. It isn't just the processor that is important, it is all the APIs, all the programming hooks that allow the program to communicate with the operating system. Windows and MacOS share almost zero of these, so you can kiss goodbye the idea of ditching Windows for MacOS and expecting all your programs to keep running. This is what people seem to forget.
Don't Wait For A MacIntel
Apple computer: Is Piracy the Pathway to Profits?