Hmm. Well let's see. I do think we are headed to the post PC world. I will not be surprised if iOS devices in use reach a billion by the end of the decade. Actually, I will be surprised if that is not the case well before the end of the decade. Also, while PCs will still be the more numerous, the vast majority of those PCs will be work related with limited uses focused on the job at which they are installed.
My assumption is that basically EVERY computer science major in school kind of wants to be the guy who programs the next Snapchat. Well maybe the guys who want to make the next Call of Duty don't. But lots of folks do want to make the next great App. That App will get created first for iOS. If Swift is the faster way to get the App shipped, I think folks will gravitate to it. Once the App is shipped on iOS and has become a success, the port to Android is just a process. It might be harder because you started with Swift, but I assume it is still going to be very doable.
----------
Well programming is very hard, so I suspect it constrains App development.
The "developers" numbers are just filled with folks who want access to the beta program. I'm a developer just so I could use iOS 7 when it was in beta. It cost $100 and it looked cool showing off the new OS for a few months.
But making programming easier should lead to higher quality. I assume application development is a constant dialogue between the idea folks and the programming folks with one saying "can you do this?" and the other saying "no, I can't program that (not within the time I have to implement or the resources of the devices), but I can program something close to it." A faster programming language will result in less no's, right?
Good point! Easier can lead to better apps. I remember one upgrade years ago that basically changed the whole language I was using. It made it easier to develop complex apps but the language became more profession (harder to learn)...
It did in fact lead to better programs.
Those that jump into Swift will have a bit of a lag, then catch up and maybe pass ObjC development times (able to make apps quicker) However, you have to look deeper...
The language is less of a factor than the vision of the designer. We lean more on design than on coding. Even coding is less of a factor as Apple (and others) provide more and more code and/or APIs.
Example, how many coders directly speak to the hardware like the camera, vs how many just use Apples interface to the camera?
Compare this to the PC, some would speak directly to the hard drive to do data recovery, others just use code to spit out reports.
The need to access the camera beyond what Apple gives you is usually small. In fact, if you don't use Apples APIs to access sensors, you risk being shut out in the next upgrade.
My point is that learning the APIs is a bigger and bigger part of developing most apps, and that means that the language is less of a factor and knowing the APIs is more of a factor.