I'm still pondering whether Google's 20% time constitutes a new kind of effective R&D for them. Here's a story that Microsoft are also thinking about how to revamp (and accelerate / diversify / distribute?) their R&D.