Obviously the iCult popularised the app concept and so has more members so I can understand why they are out in front. Still, there's good reasons for doing Android development:
- Lack of competition for titles with high production values.
- Higher specification devices (Droid, Incredible, Nexus One, X10, Legend, Desire and probably others out perform the iPhone 3GS).
- Develop your way without having to plant one on Jobs' rear end and hope he doesn't revoke your privileges (see Unity Engine!).
I wonder if there's any assumption of Android = Linux = People Don't Pay For Software. Hideously wrong - although trying to sell games on PC Linux has never been popular, Android is a general consumer platform and not a geek OS. There are legitimate reasons, mainly regarding fragmentation:
- There's 3 different resolutions on that list of higher spec phones, which isn't as convenient, plus others (Tattoo's QVGA for example).
- Not all Android phones outperform the iPhone. G2/Hero, Tattoo, G1/Dream, Magic and others.
- There's all kinds of different Android versions in use. 1.5, 1.6 and 2.1 dominate and are quite different.
- Some have hardware keyboards, some don't. Some have trackballs/optical eyes, some don't. Some have multitouch, some don't.
- Limited storage space on the phone, so large apps need to download assets when run, meaning you need the network infrastructure, bandwidth and error checking. Sticking the whole thing on the Market would be a lot handier.
All of these mean you've either got to handle numerous different combinations or cut off potential customers. None are a deal breaker, but none are ideal, especially for a bedroom coder who can't buy loads of phones for testing on each. Google really need to let apps be installed on the SD card (or at least let apps downloaded from the market store 99% of themselves on the SD card and the program logic on the phone if they want to pretend they're stopping piracy...)
I'm confident Android will take over eventually and the iPhone will be a niche product like Apple's desktops.
For now, I think creating the logic and art assets for an iPhone game would mean you could port to Android quite easily. The code would need rewritten and retested and neither is a trivial task, but the design, logic and assets are done. The same goes for porting to lower priority OSs like Symbian/MS Phone/other.