(This post got long. tl;dr: Microsoft's approach with WP7 has a significant impedance mismatch with the carriers & device manufacturers. Will end-user dissat with Android be strong enough to overcome this impedance mismatch...that is the question.)
The fact that Windows Phone has, thus far, avoided fragmentation (almost every WP7 device from every manufacture r & carrier automatically got updated to WP7.5 "Mango" this fall) actually points to one of the core reasons:
_The device manufacturers, mobile operators, OS providers, and end users operate in an overly complex virtuous cycle_
A virtuous cycle is one where each side of the market both gives and receives positive value from the other sides. So much positive value is exchanged, with low friction, that the cycle grows and grows, like a snowball rolling down hill. The more sides to the market that exist, the more complex the system.
In the mobile device space the four primary sides of the market are not actually aligned very well. In fact, there is such deep misalignment that there is great instability. Android has succeeded by capitalizing on that misalignment. Windows Phone is attempting a different strategy...
Carriers: Own the customer. Own billing. Own Sales. Own the physical pipe. They hate being just a fat dumb pipe, but their capex structure means they will never be anything but a fat dumb pipe.
Device Mfgs: Own the hardware. Own the industrial design. They hate not owning the customer. But their HW bias (and manufacturing capex structure) prevent them from breaking out of this.
OS providers: Own the core of the customer experience. Own most real innovation. They hate not owning the customer. Their core business models (search, desktop/server OS, office, ...), as well as the fact they can't build HW, means they are always at the mercy of some middleman between them and the customer.
Users: Own the disposable income. They don't know what they hate. All they know is they buy phone service from mobile carriers and/or buy a phone from a carrier. They love speeds & feeds and will generally buy anything they are told to by television ads and RSPs (Retail Sales Professionals).
Note that Apple circumvented this by cutting the device manufacturer out and used that fact to force the carriers into being even more of a fat dumb pipe. Topic for another day, but my belief is over time this strategy will start to deteriorate for Apple.
Android has been wildly successful because it was built to reduce friction between all sides of the market. It 'bows down' to the device manufactures AND the carriers. It enabled device manufactures to do what they do best (build lots of devices). It enabled carriers to do what they do best (market lots of devices). It enabled users tons of choice. My hypothesis is that it also enables too much fragmentation that will eventually drive end users nuts.
Windows Phone has taken a different approach. It raises it's middle finger at both the device manufactures and carriers. It says "here's they hardware spec you shalt use". And it says "Here's how it will be updated" (to the carriers).
Thus both of those sides of the market are _reluctant_. Especially the carriers.
This is why, despite being a superior PRODUCT to Android, Windows Phone has not sold as well.
The question in my mind is whether Microsoft's continued investment in WP and close partnership with device manufactures such as Nokia will eventually enable a breakthrough here. I know that MS can be very persistent & patient; it's been so in the past. We will see.