I love how people try to convince themselves and their mothers that the Android OS fragmentation is a non-issue due to the users not caring. Have any of you ever tried to develop an app, ever? Have any of you any idea how developers and people actually trying to make a living out of development are looking at this whole fragmentation?
Because that's what really matters in the end, because whenever the fragmentation is slowing things down for developers it's not only affecting the users that doesn't care, it's affecting those who live in the cutting edge, root their devices and whatnot to keep their phones with the latest and greatest.
I can fully agree that cheap Android phones, sold for less than a meal, used by users who love to just use their phone as they are, not updating, installing, doing nothing at all with their phones other than make phone calls, send a few text messages and use the built-in camera application and web browser like two times a month don't need be always up-to-date and running cutting edge software at all times.
But this fragments the whole Android market in such a way that it makes things much more difficult for the developers, and it makes the whole "so many more use Android compared to iOS and Windows Phone these days" argument rather lacklustre because whom many do actually use competent Android phones? If these users with their cheap, not up-to-date phones are not to be accounted for in-terms of actual fragmentation of the Android ecosystem why should they even count for the overall Android market share at all?
For a developer seeking to ear money out of a platform, this is basically what builds a ecosystem to begin without, if there is no developer, if there is no development and no apps coming when they are relevant then there isn't much of a ecosystem by today's standards so for a developer it boils downs to two options:
#1: Build applications towards the majority of users, and considering so many are still running old version of Android that would include developing towards and optimise for older hardware and older software.
#2: Ignore the users on old hardware and old software, which means you can slice off over 30% of the Android userbase and all of a sudden Android didn't really have any higher market share relevant for developers to begin with.
Most developers goes for the #1: option, they want to include everyone in order to maximise their potential userbase making their applications more relevant and more likely to be adopted and provide the developers with a income, which is what this is all about for developers.
What this does is making "cutting-edge" hardware and software on Android phones barely utilised at all. This is why about every iOS app feels as fluid, featuring at least the same amount of detail and feature as their Android counterpart even though Android phones tend to boost more cores, higher frequencies and whatnot. There is no denying that mobile gaming is much larger on iOS devices compared to Android devices, there are simply more games, games featuring more advanced and better graphics all this running on devices like the iPhone 4 and 4S which are not remotely close to the performance and hardware of newer Android phones like the Samsung Galaxy S4, HTC One etc.. And the one could ask what is the point of such hardware if it's not being utilised and optimised for by developers?
There are some exceptions, but there is clearly much better optimisation towards hardware on the iOS platform and this will always be the case when Android phones keeps coming in such a verity of flavours, some featuring 800MHz single-core SoC, others featuring 2.0GHz quad-core Soc, some featuring 3.0" low-resolution displays, others featuring 5.0"+ 1080P displays, some running Android 2.X, others running Android 4.X. It's just too many devices, too much hardware and software for developers to take into consideration making optimal optimisation towards all devices utterly impossible, good scaling and whatnot can only lead you half way there.