If Google had sent that much to developers they'd be saying something. 90% of the data points we see are of the form:
My app made X on Android and 5-10X on iOS
The reason those "data points" are skewed is psychology. No one is inspired to write a vitriolic blog post whining that their Android app only makes 73% as much as the iOS version. Nor is the iOS-centric HN echo chamber well equipped to give you a random sample of these things. And, frankly, lots of those bigger numbers are simply old. Android's market share is growing rapidly. Three years ago, I think 10x was probably correct.
And 2.3 was 2.6 in January so for a really bad back of the envelope calculation we can guess 2.3 will be 1.0 sometime before 2017.
I think there are lots of advantages and good reasons to target both, but as long as the majority of the revenue comes from iOS, it makes sense to target iOS first.
Also, lots can happen between now and 2017. 4 years ago nokia and blackberry were still competing. iOS or Android could in theory be small players in 4 years. (Note that I don't think that will be the case, but it's possible.)
> Nor is the iOS-centric HN echo chamber...
Read the comments here. The significant majority are pro-Android. The whole 'plucky little Android' schtick is extremely tiresome and frankly devalues your point. The "psychology" that you cite is basically confirmation bias, so beautifully illustrated by your comment.
Whilst Android is making gains is some markets, so Apple is in others. Apple is up before a hardware update in the UK, France and US, taking share away from Android.
Of the data points that in your circle you see, you mean? I've seen anecdotes recently of publishers making much more on the Android market. That's the thing about anecdotes.
The most recent stats are that they are closing very fast.
But it is incredibly variable. How are they marketed? Do they pander to the demographics? Do they take advantage of the platforms? Are they crowded segments on each platform? How does it compare to the incumbents in those segments? Etc. It is impossible to separate all of those, which is why such comparisons are usually bunk.