Money? An enormous and dedicated user base? Users who expect high quality UX and are willing to pay for apps that provide it?
edit: Wait, I forgot a few: a very limited number of hardware platforms to target? The ability to target at most, two OS revisions and cover the vast majority of users?
It is, in many ways, an easier platform to build high quality software for. On top of that, the users there give a damn about the quality of the software (and not some ephemeral quality like "openness").
> Develop for open systems instead. They will win in the end.
"In the end" is awfully open ended, and really there's very little to demonstrate that open systems always win in the end. Exceedingly closed systems sometimes fail fast due to lack of adoption, but that doesn't really hold in this case. Apple consistently releases hardware and software that, despite being closed, many users perceive as "best-in-class" (it does not matter whether you agree that they are -- the fact remains that many users believe that to be the case). Developers back that up by building best-in-class software for the closed platform because some of them also believe it to be best-in-class. As long as you've got that feedback cycle in place, I don't believe a case can be made for openness even being a substantial factor in the shape of the market.