Folks, I think a ton of you just aren't getting the major possibility here. This is Apple's chance to deal its first substantial blow to the Microsoft/Windows hegemony since sliding into second place 20-years ago.
MS chose to not build a modern, clean-slate OS, and, is hemstrapping Longhorn with all of the dead weight required to maintain backwards compatibility with four or more generations of prior Windows releases. They made the business decision to not make a clean break at this time.
Tiger is already everything a clean-break "Longhorn" should have been. And, if Apple simply specifies a minimum system configuration on which the new "Intel" Tiger will run, it limits the overhead on the company of supporting a bazillion antique drivers and APIs. And, it puts Tiger up there as a surprise, solid alternative to Windows for new hardware buyers.
Apple will not attempt to support parallel ports, floppy drives, serial ports, etc. with this move. It will start from the same hardware assumptions that now underpin the current Mac lineup, and will require a certain base Intel hardware configuration in order to run the revised Tiger.
Is any of this work needed on Tiger (or major apps) trivial? No. But, if Apple's been developing toward this transition for some extended period of time, my guess is that all of the iffy parts have long since been worked out.
Personally, I want one of those Voodoo Eden air-cooled dual Opteron towers with OS X running natively. From a profitability standpoint, that hypothetical sale of one copy of Tiger would earn Apple about the same as if I bought a Mac mini, or an entry level iBook or eMac.
Nothing Apple can do will result in a sudden, large migration of corporate desktops to OS X. But, the consumer, small office, education, and (don't forget) the enterprise spaces are not as tied to Microsoft. Huge sales gains can be had in these spaces with an x86 compatible OS X.
