Given WP7's pretty disappointing launch so far, Nokia is customer #1 at MS. They can dictate the direction of the platform, and as the de facto (and one of the only) WP7 OEMs, the WP7 user experience will become synonymous with the Nokia user experience. They are in essence taking HTC's place back in the old Windows Mobile days - as one of the only actual supporters of the platform, they got front-row seats.
On Android, Nokia would've had no say in anything - being in the company of juggernauts like Moto and Samsung. They'd have trouble differentiating their UX from anyone else's, resulting in the commodification that Samsung, LG, and Moto are facing now. They'd be stuck in the same position as all the OEMs now: a UX they don't control, branding diluted across your competitors, and relegated to shipping crappy UI tack-ons in a desperate attempt to differentiate your Android offering from someone else's Android offering.
The WP7 launch hardware has been disappointing so far - build quality is middling, and resembles far too much like the hordes of Android phones out there. MS would do wise to tie WP7 tightly to Nokia - who have in the past created some of the best hardware in the entire industry (design and build quality inclusive).
The jury is still out but the general sentiment seems to be that Android is better software than WP7 and as you insinuated the "hordes of Android phones out there" aren't particularly well built, certainly not to the level of the iPhone. If Nokia is as good at hardware as they are being given credit for wouldn't they want to build premier hardware for the Android platform? It's not like MS will not license WP7 to all the other OEMs that are on Android right now and I haven't seen much evidence that Nokia is actually good at UX so it's unclear what they could add to WP7 that would be of so much value. After all "taking HTC's place back in the old Windows Mobile days" seems like a pretty scary image of the future.
If all they do really well is hardware, and both platforms are available to all OEMs, why not go for the more successful one, especially considering that WP7 is proprietary so MS can decide to play hardball with them at any point.
My own opinion of Nokia is that their classic line of phones had by far the best software, and while initially it had better hardware, that lead narrowed considerably pretty quickly. Yet that software was very good because it was an extremely feature limited and highly resilient phone OS. Nothing modern is as good at being a cellphone as a classic Nokia. And as the iPhone conclusively demonstrated, that doesn't matter these days. Symbian was a disaster and Meego seems promising but behind. They really lost a great catch in Palm, and I'm not really sure they'll be able to differentiate themselves as a hardware builder on any platform for that matter.
My weekend coach suggestion would be to take a leaf from the iPhone playbook and release only one phone every year or more, that has the best possible industrial design you can muster and that gets timely updates to a vanilla version of Android, properly tested and integrated. If you want a low-end model just tweak the previous year's high-end model and sell it cheaper (think PSOne).
On Android, Nokia would've had no say in anything - being in the company of juggernauts like Moto and Samsung. They'd have trouble differentiating their UX from anyone else's, resulting in the commodification that Samsung, LG, and Moto are facing now. They'd be stuck in the same position as all the OEMs now: a UX they don't control, branding diluted across your competitors, and relegated to shipping crappy UI tack-ons in a desperate attempt to differentiate your Android offering from someone else's Android offering.
The WP7 launch hardware has been disappointing so far - build quality is middling, and resembles far too much like the hordes of Android phones out there. MS would do wise to tie WP7 tightly to Nokia - who have in the past created some of the best hardware in the entire industry (design and build quality inclusive).