So having worked for Nokia previously (actually Symbian and then we got bought) I think the basic problem that they are trying to resolve is the sheer amount of dead weight the have in the organisaiton. There is a reason they have the most expensive and least productive R&D operation on the planet and that is because they get so little out of each employee. Most employees are jobsworths simply doing the minimum they can get away with without being fired.
Think about it, all those employees, and they couldn't be co-ordinated to create a winning platform. Whilst there is definitely a degree of management failure, there is also a severe lack of personal responsibility and accountability at the lowest levels of the organisation. Moreover, with the long running drip feed of redundancies over the last few years, most of the talented, motivated engineers have left.
The net result is that this has left a big soup of shit, that they call an R&D operation. I think that Elop has done the right thing by clearing the decks. Obviously a shame for some of the people, but life moves on. Once most of the existing people have gone, and legal obligations with regards to re-hiring roles you've made redundant have passed, I think Nokia will start re-building their R&D from the ground up to be more dynamic and more responsive to the market.
Jury is still out on weather the MS deal is the right thing, and it certainly has the smell of Elop being exposed to a single ecosystem for so long that he wasn't really able to properly evaluate alternatives, but it is probably worth a try in the face of Android genericism. Although given Microsoft's double-take on Silverlight recently it's already starting to look a bit wonky.