I actually think WP7 will fail much worse than Vista. Vista was a bit sluggish but it run the old applications. WP7 can't, and that will be fatal. All the Windows Mobile users will move to Android where their apps already work. People who already have an Android or iOS device are very unlikely to switch to WP7. All the ISVs will end up on Android and/or iOS because it's easier to port an app to a platform where you can use C/C++ and native code than one where the whole thing needs to be in C# and Silverlight or XNA. Even Angry Birds needs a C physics library. In fact even if Microsoft allow C and native code I doubt the ISVs that used to support Windows Mobile will come back because the platforms already bad market share is dropping quickly.
E.g. Pleco - a Chinese dictionary - moved to iOS and (soon) to Android. They've dropped Windows Mobile and won't ever support WP7. When they dropped Windows Mobile the iOS version was outselling WinMo 10:1. They have core code in C/C++ which they can run on both iOS and Android (also on WinMo). No chance of it working on WP7 without rewriting in C#. And no chance of getting their handwriting and OCR libraries from third parties ported either.
Opera have dropped Windows Mobile and won't support WP7. Once again they have C/C++ code with a few third party libraries in native ARM. It would be almost impossible to port to WP7 and even if they did Microsoft have apparently said they won't allow alternative browsers in their app store.
In a sense WP7 is more like a console than a phone. Worse actually since XBoxes support native code as far as I know. Maybe they'll pick up games from the XBox ecosystem but I don't think that will make up for not having things like Opera and Pleco though. They've apparently offered Adobe the possibility of native code to get Flash ported and possibly will do the same for titles like Angry Birds. Still that's not really enough - Adobe haven't announced a ship date and Roxio, the Angry Birds publisher, have publicly contradicted Microsoft when Microsoft implied they had committed to porting. I.e. handing out native code passes for key applications is not enough to get people to support a platform which is obviously doomed.
Picture Vista with no back compatibility following on from XP which had 1/3 the market share of OSX. Imagine that all the software already worked on iOS. That's the situation WP7 is in - it's actually easier to run the apps you used on Windows Mobile on Android than on WP7. Even the IHVs like HTC prefer Android because it's free to them and there are no limits on things like the Sense UI. WP7 has ridiculous limits on how much value they can add and they need to rewrite all their WinMo software in C# to make it work.
I think the market share will drop rapidly and Microsoft will kill it. Just like Kin and Zune, both of which used the same software.