My impression is that there's never been a particularly good relationship between the product groups at Microsoft and the more long-term focused researchy group

My impression is that there's never been a particularly good relationship between the product groups at Microsoft and the more long-term focused researchy groups (not necessarily limited to MS Research proper?). A lot of "employees in the trenches" of the product groups seem to perceive the researchy guys as useless, impractical, a waste of money, etc. Ozzie's concrete responsibility at MSFT, so far as I can tell, was running a bunch of medium-term focused groups to try and bridge that divide. They seem to have been generally regarded as unsuccessful in that respect so it's not too surprising that he would eventually quit/be pushed out/whatever.