Apple is in the position Sun Microsystems was 20 years ago. They supplied finished, working solutions, which were really "just" high-quality versions of what the other guys did. Eventually, services got so widespread (Lintel) that customers got less picky and Sun's peculiar quality didn't matter anymore.
However, that won't be how Apple will die out (if they do...), because there is already a huge market of people with lesser quality standards than Apple (Android). Instead, Apple's threat is another company who takes the time to give extraordinary quality for novice users. This isn't happening right now, b/c everyone is fighting for the disruptive (low) end of the market.
So, Sun could have extended their business by diversifying to serve specific high-volume customer markets. Java was an epic half-step. They gave away their ability to add value to customers for nearly free, while focusing ever harder on custom unix hardware and software (UltraSPARC!).
If Apple wants to avoid a Sun-like fate, they'll need to properly diversify beyond custom hw and sw and extend into the spaces of large customers where they can add value. You see _some_ of this in their supporting multimedia Bitflingers, but they'll need to incur much deeper and much more specifically into the multimedia and small office spaces.
Perhaps the Cyanogen corporation could morph into a "luxury" distro?