The Waiting Game
It’s not often that I can find fault with John Gruber of Daring Fireball. However, I think he may be downplaying some things in his latest entry Bombs Away. Considering all computer users are always looking towards their next purchase, the issue of when to buy is always paramount. He said:
This transition period is going to be hard on Apple, not hard on Mac users. If you?ve just purchased a machine recently, or need to buy one soon, you?re no worse off than you would have been if Apple had remained committed to producing new PowerPC hardware ? today?s machines would have been obsoleted by even-better machines next year no matter what processors they contained. And no one is abandoning PowerPC software development. I see no reason to expect Intel-only Mac software in the near future. Universal Binaries take full advantage of both Intel and PowerPC Macs. That?s worth repeating: Universal Binaries take full advantage of both Intel and PowerPC Macs.
This smacks a little bit of cheerleading. There’s nothing wrong with being optimistic, but when you’re over the top, you lose credibility. Luckily, I have none to begin with, so I don’t have that to worry about any of that. However, John does, so to see this kind of statement out of him is a little disconcerting. Let me highlight the portion that isn’t quite right in my judgment.
Universal Binaries take full advantage of both Intel and PowerPC Macs.
Bolding should drive my point home a little more. Universal binaries will run on both platforms. They will not necessarily take full advantage of those platforms. In order to take full advantage of both processor types, processor specific optimizations have to be implemented in both. Seeing as x86 is the future of the Mac, I fully expect developers to begin integrating SSE code when applicable. I’m sure Adobe is gearing up to transfer all that optimized code from their Windows version to the Mac version. The real concern is whether Altivec code will be developed anymore. Considering the limited shelf-life of PowerPC Macs, I don’t think so. Spending that extra development time for code that will soon be obsolete would be foolish because of a very low return on investment.
Then there’s his other point.
And no one is abandoning PowerPC software development. I see no reason to expect Intel-only Mac software in the near future.
For most classes of software I would expect this to be true. However, I foresee that Mac games (should companies still find profitability in it even after Macs gain the ability to dual-boot Windows) could very well start doing this next year. It wasn’t terribly long before games started requiring G5s. Add in absolutely astounding integer operation performance on Intel’s chips, and we’re looking at a recipe for Intel-only games.
It’s for these two reasons, the lack of PowerPC optimization or support at all, is enough to keep me from buying until I can get an Intel Mac to my liking. The same goes for everyone in my family and anyone that seeks my advice for future Mac purchases. There are extenuating circumstances for some, but for the majority of users, there is no reason to be so impatient as to not wait a year (which is not all that long). By waiting just a little bit, you can avoid getting locked out and ensure a long life for your new Mac.

Excellent points in your piece.
Another thing to consider is the very good prospect of fast Windows emulators for Intel-based Macs (not like Virtual PC now).
A good thing, right?
Wrong.
If Intel-based Macs can run Windows software at acceptable speeds, this leaves open the very real possibility that a lot of developers will simply have their Intel Mac customers run the Windows versions and they will drop the Mac versions or let them languish.
This will save them enormous amounts of money, time, personnel and effort. It is an extremely attractive proposition to many developers of consolidating all development under one platform instead of two.
Writing the same app for two different platforms is a hassle. Many would jump at the chance of a good alternative. And running Windows software on an Intel Mac at near-PC speeds is a good alternative.
This could cause a mass exodus in the Mac software market. And if there’s no Mac software, the platform will die. We could witness the end of the Mac as an independent platform.