I've always felt that Apple's problem is they hit on a successful, good formula, and then ride it into the ground and well past its sell-by date. Original Macintosh was something new and interesting, but they were still making the Mac Classic into the early 90s which was fundamentally similar and did poorly outside education (trying to sell a non-color 8mhz computer when PCs were at 33 and 66mhz? Really?). The early color Macs were interesting, but they got bogged down in the confusing product lines and corner-cutting such as the 5200/5300/6200/6300 models. The classic Mac OS was great for 1984, but it was already out of date by the early 90s much less Mac OS 9 running into the early 2000s.
If they were better able to recognize when a market was done and when they needed to bring out new products, they'd probably do well swinging back and forth. They've been on a big upswing since the iProducts and Mac OS X were introduced, but it can't last forever. Granted it's anecdotal and hardly indicative of anything but I have several Apple fan friends who bought Galaxy S3s because they were unhappy with the iPhone 5. It's probably still got several years left but it'll be interesting to see if Apple leverages their iOS lead and fandom into a new world, or we end up in a world of Performas and Centrises again.
Not that I care personally, I'll keep my Android and Linux thanks. More an academic interest as a computer guy.