As the second-generation Android devices debuted, there were serious questions about the state of the platform. Unlike the release of the HTC G1 (also known as the "Dream"), which was for a short time the only Android handset available, more phones and more versions of the Android OS meant that manufacturers and wireless carriers had to make decisions about what which hardware and software to support.
The G1 was almost completely un-branded as an HTC phone. Instead, HTC gave over control of the UI to Google to ship a completely "stock" version of Android OS 1.0 (later given an over-the-air update to 1.6) and carried by T-Mobile. But Google had been clear from the inception of the Android OS: Its intended goal was to have multiple hardware configurations, running multiple versions of the OS, supported by multiple carriers simultaneously.
While this seemed like a bold move at the time, especially in comparison with the Apple's iPhone 3G, which had been released just months prior to a very warm reception from the press and consumers alike. Apple's platform seemed unbeatable in part because it was a monolithic, closed ecosystem. Sure, Apple had capitulated to demands for user-created apps, but by the time the iPhone 3G was on shelves, it was clear Apple had the consumer smartphone platform to beat.
In retrospect, however, Google's intent to deliver multiple Android experiences across multiple hardware, software, and carrier configurations seems to have proven a wise choice. Unlike Apple, Google has it's fingers in many pies, and the ability to leverage services like Gmail and Google Voice, gave them an early edge in providing unique mobile experiences untethered to the desktop the way iPhone was (and still is to some degree) wedded to iTunes.
Google developed a system that gave them the freedom to support some premium services on the devices that could use the effectively, and a more basic user experience on those devices that could not. The most recent example was the (botched) release of the Nexus One, which is still the (officially) the only Android handset to support the 2.2 ("Froyo") version of the operating system. With 2.2 came full support for Adobe's Flash platform, still unavailable on lower-powered Android devices, as well as the iPhone.
Which brings us to the state of Android OS segmentation today: Android OS versions 1.5 and 1.6 make up slightly less than 50% on Android devices, according to a GigaOm report, while versions 2.01 and 2.1 have steadily risen (most notably with the release of the Motorola Droid series of handsets).
The Nexus One, Google sold directly to customers, without cellular service, did not fare very well. As the only device to officially support Android OS 2.2, and being that is no longer available for purchase, version 2.2 of Android occupies a tiny margin of slightly more than 3% of the Android OS install base.
Surely we'll see more Android OS version 2.1 and 2.2 handsets, but the real question is if we will see more version 1.6 handsets released in the coming year. I would wager we may, and cellular service providers will use the now low-cost devices in place of what would have otherwise filled their feature-phone lineups.
The higher-end Android OS devices will continue to be billed (both figuratively and literally) as premium smartphones. The wrinkle is that cellular providers, already buckling under skyrocketing bandwidth usage (or so they claim) may not have room on their networks for a host of more Android OS 1.6 devices. Will we see Android "dumbphones"? It's probably unlikely, but a creative solution will be needed.