"The problem with Android, IMO, is that the entire ecosystem composing it and much of what surrounds it is entirely insular, and to no great benefit.
It shares no common libraries or interfaces with what you find in most Linux distributions. It uses a unique libc that no other distribution uses. It uses a file system layout that is not found anywhere else. Its GUI rendering subsystem is completely unique and incompatible with all others.
The end result is that changes to Android stay within the Android system and do not benefit open source projects outside of it. And projects outside of it require heavy rewrites to work, at all, on Android. Not to mention that Android has no real repository type system, so you're left trading .apk files and latching on to the market, which is only available on the default builds of some devices and not at all on others.
Maemo was developed with that compatibility in mind, and is a large part of the reason I bought it. It was most of what the OpenMoko Freerunner tried to be, and MeeGo only improved the openness aspect of it. MeeGo allowed mobile devices to retain continuity with the rest of the open source ecosystem you find in most desktop Linux systems, thus changes and improvements to both ends benefits everyone. In addition, it removed the non-device-specific closed bits and created a platform independent of any one handset vendor.
Android leaves you a second (or more likely, third) class citizen in this effort, as the AOSP does not, last I checked, flow upstream into the Android core and the AOSP only receives the latest changes to Android after it's been delivered to device manufacturers (see Honeycomb and Motorola.)
So this is very much a Microsoft victory against Open Source, if not Free Software, projects in the mobile space. And Android is not a way forward that is very fair to end users and non-corporate developers."