I’ve read them, and I don’t see the big deal. The former license appears to be quite MIT-like, with the addition of a couple of explicit patent clauses (contributors agree not to sue you over patents, unless you sue them). The latter is much the same, except with the additional clause that you have to pass on the licensed source code to recipients of binary code. Neither one expects the user of the code to release any of their own additions under an open source license, and the former doesn’t even require you to distribute the original source code.
In terms of Microsoft using these licenses, they seem fine to me. If I used code released under these licenses, I’d feel comfortable. They aren’t, however, licenses that generally encourage community participation I’d say - the lack of any clauses that say that additions or changes must be published (except for individual file changes in the reciprocal license) means there isn’t a huge incentive to contribute to a project using these licenses - after all, other people could be hacking away changing it / adding to it and keeping that work for themselves alone, so why should you contribute back? You might anyway, but the incentive is definitely reduced. But then, Microsoft doesn’t really ‘do’ open community development so I guess they don’t really care about that aspect, I’m sure it’s more of an open publisher / subscriber model than a collaborative one they’re looking at. Maybe, if Steve Ballmer really does want all open source development to occur on windows (as if), the stance may alter over time and a more ‘share-alike’ license might appear.
To me these licenses do deserve to be called open source (although obviously not ‘free’ as promoted by the GPL and variants), so the OSI approval is a good thing. Some people won’t like it just because it’s Microsoft, but to my mind you should just take a license on its merits. I’d much rather Microsoft used these licenses than some horrible tome-like EULA any day.