1. Scuttle the Nokia Android fork
It's all but certain that Nokia will announce next week new smartphones running a forked version of Android that uses the Android Open Source Platform (AOSP) and a combination of Nokia and third-party services to replace the Google services found on a "real" Android device. But AOSP is an iffy platform on which to base your future. Even though Microsoft's acquisition of Nokia won't close until March, it should block this dumb move now.
I understand where this is likely coming from: Nokia sold 30 million Lumias in 2013, and Microsoft recently says it needs to sell 50 million a year to break even. That ain't gonna happen. Some salesperson likely said, "Hey, lots of people buy Android, so let's do our version of that, like Amazon did to create the Kindle Fire." This kind of defensive market reaction is what happens when salespeople run corporate strategies, and it's a key reason why Microsoft is in its current mess.
Perhaps someone in Microsoft should recall its spectacular failure with the Kin. There once was a briefly hip mobile device called the Danger, and Microsoft bought it as it flailed in its response to the iPhone and Android. Windows Phone still wasn't ready, so this was another defensive response. It was poorly designed, confused the market, didn't sell, cost Microsoft a lot of money, and revealed a big crack in Microsoft's reputational armor. A Nokia Android fork would likely have the same result. After all, it's not like Nokia has a clue on how to succeed in the smartphone market, either.
2. Kill Windows RT
Microsoft's tablet strategy has been even worse than its smartphone strategy. Windows 8 tablets have the Metro touch environment but also run the full Windows Desktop that works terribly with touch and is hard to read on a tablet-sized screen. There are few Metro apps, so you're really stuck with traditional Windows -- which is simply painful. Windows RT tablets run only Metro apps -- well, plus an embedded version of Office and Internet Explorer that mostly mimics the Windows Desktop version, with all the usability issues that causes. That's painful, too. No surprise that sales of both Windows 8 and RT tablets are too tiny to show up in most analyst surveys.
Windows RT was supposed to be Microsoft's iOS, meaning a mobile-specific operating system that would have synergies with the Windows mothership, as iOS does with OS X. But the Desktop-derived version of Office essentially meant no native RT apps were to be had. Contrast that with the iPad, which has several very capable office productivity suites designed for its touch environment, versus none for Windows. Android too has a decent set of Office-compatible apps. Microsoft's positioning of Windows 8 tablets as the "you get everything" device also ensured no serious effort would be spent on Metro apps, whether by Microsoft or other developers. Meanwhile, iOS is chock-full of highly capable apps; Android now has a pretty decent set as well.
Windows RT is a distraction that needs to go away.