Feature driven competition, and the Android problem

Tim Bray just posted a nice blog piece about his new phone: a pre-production Nexus Galaxy running Android 4.0 (code-named “Ice Cream Sandwich”). There are some really nice features in this release, which is, of course, what we’ve come to expect. These days, competition in many markets is driven by features, and less attention is given to price, performance, quality, and customer service.
But will I ever get to use these features? Here’s the comment I posted on Tim’s piece:

Yes, Tim, it all looks very nice. But I’m still waiting for AT&T to get around to updating my Samsung Infuse 4G from Froyo to Gingerbread (promised in August, already shipping in Canada). I have no idea if AT&T and Samsung will ever put Ice Cream Sandwich on the Infuse, let alone when. I read your account of the new features, but I have no idea whether it’s relevant to me.
For me, this is the biggest bug in the Android business model compared with iOS: it’s completely unpredictable. All of the players – Google, handset makers and carriers – contribute to the mess. And so I’m not surprised that so many apps are so unstable: the test matrix is ridiculously big.
In contrast, when an iOS release comes out, I know exactly what it will run on, and which features will be available on my device. Moreover I can install it immediately.
This summer, I decided to try life outside the “walled garden” and replaced my iPhone 4 with the best Android device then available. I have to report that so far, life outside the wall sucks. This is a shame. I guess I could hack it, but great products shouldn’t need hacking….