The Kindle Fire really shouldn't be counted as an Android device. What's next, a press release stating Linux hits an all-time high market share, running on devices such as the Mac, iPhone, all Android phones and the Raspberry Pi?
Also, customers would probably not hold their Kindle Fires right now if it weren't for Android. Developing an OS is an awful lot of work. And while Amazon's intentions are to fork Android, I'm pretty sure they'll come to their senses and just treat their version as a branch.
They'll still insist on having their own curated app store and their own browser of course, but it will be an Android device nonetheless.
The issue with the Fire is that it's forked, that there may be future compatibility issues, that it will never officially become an ICS device and that those things may hold Android back in may ways (for instance app developers always having to work to a low baseline to access the Fire users as a potential market).
So yes it's an Android tablet but it's one that may be a mixed blessing for Android and deserves some subtlety in the distinction.
I'm seeing this "fork" more like that of Debian and Ubuntu: different software bundled by default, different repository, different philosophies and policies, different management, overlapping user-base, but for practical purposes, they aren't that different.
But developers are targeting the phones first, tablets second, because a phone is always in your pocket, always online and everybody needs a phone. In December Google reported 700,000 activations per day. In 2010 alone there were 67 millions of Android devices sold, with the total number of mobile phones sold being measured in the hundreds of millions.
So Kindle Fire does great as far as tablets go, but it can't compete in numbers with Android phones and Android phones are targeted first, unless you're talking about stuff like interactive eBooks that really need the bigger screen.
I'm also seeing this happening on iPhones and iPads. Unless an app really needs the bigger screen, then the iPad is a second class citizen to the iPhone.
That's a good point about phones leading the way. I'm guessing that in large part that's because the number of phones (iPhones and Android) dwarfs the number of tablets. I'm always slightly surprised at the number of apps that don't work well on the iPad but it makes sense that you wouldn't put your energy there, at least not in the first instance.
But for those applications that do want to develop a tablet app, it's still a big call to exclude the single biggest selling Android tablet out there (which is probably the Fire).
1. Amazon and B&N are in a dogfight over eBooks, eReaders and (now) "reading tablets". It would be stupid for either of them to skip all of the new functionality and internal improvements Google has released.
2. Plenty of apps Amazon wants in their Appstore are also going to support (and want to use features from) ICS. While Amazon could make them jump through unnecessary hoops, that just makes their Appstore (and, by extension, the Kindle Fire) weaker.
3. Amazon itself is going to be developing applications that support ICS (e.g. Kindle, Amazon MP3, Amazon Mobile, Price Check, ...) and those apps will, in some cases, need to use ICS-specific features to be competitive. They could fork their own internal applications, I suppose, but why would they? Would they really want to risk a situation where an ICS version of an app is better than the Kindle Fire's version of the same app?