The trick to that model for Microsoft is how to get paid. They have been trying to push Win CE on a paid licensing basis. This has been the norm until Google came along. Now Microsoft has to figure out how to compete and still stay in business.
The win here for ARM is not necessarily Microsoft tablets. As pointed out above, ARM already has a solid hold in that market. The win is in server farms. A solid Windows family commitment gives ARM a whole new market segment in terms of entry into the massive server farms, where power usage is a real concern.
I've got a small, 2.5W, system running on CE5 on MIPS/RMI from MSC and the multimedia performance is great.
Regarding Windows ... currently the concept of an Operating System is slowly vanishing so the hardware platform used becomes even more irrelevant. More and more anything with a web browser is enough for the majority of users. In this case MS has to find a new target, perhaps some "internet terminals", ARM+Win?
Wish I could predict rumors. The psychology that makes ARM shares jump on the rumor that Windows might be ported to it must be really fascinating (if you're a head-shrinker). But it doesn't look like there is much choice for MS if they want Windows to continue as a viable product, given the (increasing) popularity of certain tablets.
Does anybody see red flags here with Microsoft designing their own ARM chips? In the SW arena they tend to adopt standards, add proprietary MS only extensions and then leverage all competition out of the market place. Google watch out! On other news, windows low power? you gotta be kidding. also, I've got a pioneer car entertainment system with WinCE running on ARM, and the performance is pathetic. If the audio side wasn't so good I'd trash it.