Do you know what Apple's biggest ever strategic error was prior to Mactel? The thing that drove their market share from something in the order of 9% to around 3%?
Back in 1993/4 when first Michael Spindler and then Gil Amelio were in charge at Apple, Microsoft was hyping Windows NT as the future and had Windows 4.0 in beta. NT looked better in theory than reality, but 4.0 looked quite good. Unlike Windows 3.11 it booted native -i.e. didn't run as a DOS application and it was simple enough to be reasonably solid - certainly far more so than Windows 3.11.
The only problem was that most of the PCs around couldn't boot and run 4.0, so Microsoft eventually re-embedded the win32 APIs in a DOS application, aka Windows 95, to maintain backwards compatibility with its installed base.
Apple, however, concluded sometime in late 92 or early 93 that Microsoft's new OS(es) would compete well against MacOS, result in reduced sales, and force the company to compete more on price than on ease of use and integration. As a result they de-emphasized the ease of use and integration messages in their advertising, started to use more PC components like graphics cards instead of designing and building their own, and cleared the decks for the expected reduction in sales volume by moving a lot of production capacity to third parties - out-sourcers.
When Windows 95 finally did come out, it had the opposite effect: driving up interest in Apple's GUI and creating a massive order backlog at Apple - whose out-sourcers were mainly PC builders working to PC quality standards: something Mac loyalists did not expect and would not tolerate.
In effect Apple's top people over estimated Microsoft's ability to deliver, and then set about making the consequent prediction about a decrease in their own business come true by using PC components and suppliers to dig themselves a deeper hole.
That disaster laid the ground work for NeXt's reverse take-over of Apple, the second coming of Steve Jobs, and the return of the original fully integrated Mac in the form of the iMac.
There are millions of iMacs still in use - my mother in law, for example, has one we helped her get in, I think, 98 or 99 and uses it every day for email, recipes, internet access, and some Bridge games she's fond of.
Her iMac is part of an important underground Macintosh phenonmenon: Apple may only get a bit more than 3% of new sales dollars, but it has something in the range of 12% to 15% of total users. There are three closely related reasons for this: older gear continues to work, the long product life cycle discourages early product replacement, and most Mac users have only one computer.
In the PC world new generations come along about every six months to a year. As a result the typical home user accumulates several machines while corporate evergreen policies usually aim to refresh the desktop every 18 months for power users and every 24 to 30 months for people perceived to have more limited needs.
In contrast the PPC macs now being phased out in favor of the Mactel line have had a marketing life ranging from five years for the powerbooks and G5s to two years for the mini.
One of the effects of Mactel is that this traditional longer product cycle for the Mac will be going away: a decision to use PC chipsets and architectures is also a decision to adopt the PC lifecycle.
As a result that long term Mac user phenomenon will be wiped out, shrinking the aftermarket for Mac upgrades and software in the short term, and bringing Apple's usage share down to its market share in the longer term.