I think the correct statement would be iOS has "less" fragmentation. They still have fragmentation though. Not only through the devices but also the version of iOS running on the device. It can get pretty hairy trying to maintain support for multiple iOS devices with versions of even just 4.3 and up.
Look at Google's own wallet app -- it only runs on only 6 android devices. If you want to run google wallet on your Nexus S, you'll have to run a 3rd party "hacked" version of Google Wallet.
Android development and testing really isn't all that painful, and most of the new APIs you might miss are backported and available for the more prevalent Gingerbread and now ICS.
Probably games developers have it harder on Android, but for the average app it seems just about the same to me as a developer.
There is little expectation of iOS 4 support at this point, because users who purchased an iPhone in the last 4 years don't have to use it. The same is hardly true for Gingerbread. Most of the phones that shipped with that half-baked OS will have it until they're retired into a drawer somewhere.
Every developer can safely target iOS 5. With Android it is still 2.3.
iOS is fragmentation. Android exists because 10 years ago, every mobile vendor was making his own incompatible operating system. So a developer had to make 300 different versions of his app to reach his audience. That was impossible, so there was no market for mobile apps.
Today, there are still 300 different devices, and maybe a 100 different operating systems. But because ~95 of these are based on Android, you can write a single cross-platform software that runs on all these systems. Android was build to defragment the market, and it was a huge success.
It might not work perfectly, but adapting your App to some strange behavior in some fancy Android device is still a lot less work than making a whole new app from scratch!
The only interesting OSes that you have to write everything from scratch for today are iOS and Windows Phone. These two systems are, unlike all the others, not compatible with each other and require a significant amount of extra work. And that is called fragmentation.
If you're making something that you give away for free and make money on the volume of participants somehow, then Android is probably a better bet at that point, but if that's your strategy you probably want to target everything.
I wonder how the mix of average money spent on iphone vs android apps per consumer has changed?
Or at the very least, total dollars spent on apps on each platform?
Are you optimizing for making money, or for making impact? There's no right or wrong platform (iOS or Android), it just depends on what your goals are.
We're developers right? We need to find the market with the most potential. If I'm developing for a platform where the platform holder is laser focused on maximizing their profit, but not their marketshare... it seems like I've made the wrong choice.
I see a lot of iPhone devs (not you specifically) try to justify Apple's shrinking marketshare by saying it's fine since Apple is still taking the total profit in the industry. Either they've heavily invested their earnings into AAPL, or they're trying to pull any stat they possibly can to justify their platform choice.