More constructively, I'd like to see Microsoft taking more risks and leading more. Kinnect and XBox Live are good examples of Microsoft being first in or coming up with the best thing out there. You could make a case for elements of Bing too.
So be more ambitious. Too many MS entries into markets now seem me too relying on the MS brand to prop up an otherwise ordinary product (Zune?). You need to either be better on day one or moving with such momentum that you're confident that you will be very quickly. How are you going to move with that momentum? I'd suggest less management and less marketing. Yep some of projects which don't have the current level of guidance are going to bomb out heading in totally the wrong direction, but the ones that succeed will be better than what you're producing now and get there a lot faster.
Your historic approach of watching markets develop and muscling into them using Windows and Office as leverage won't work any more. Windows isn't an effective lever in the new markets and the competitors (Apple and Google) don't shift as easily as they used to. You need to be in earlier and better.
Your developer tools are pretty good but your licensing is a quick route to insanity. I've honestly recommended Oracle in the past because for a proposal I had to put together very quickly, working out the licensing cost for a complex SQL Server implementation was simply too time consuming.
Mobile - you seem to have become obsessed with the iPhone where frankly you're ill equipped to fight. Aim at corporate, go for RIM. You're strong in corporate and if you can't do a better integration with a corporate infrastructure (which is likely based around Exchange) then you shouldn't be playing at all. Once you've got that sorted, then look at the consumer market building on that.
But also good luck. Though I don't think you'll ever be what you were historically, I honestly believe MS could be a company turning out great products and the more competition in the market the better.