Microsoft's research-and-development spending hit a record high in 2008, according to its most recent annual report. At the same time, the company's R&D spending relative to employee head count has gone down.
Not that it matters. For all Microsoft's spending on the future, it continues to focus its business on guarding the past. Yes, it builds cool (but useful?) things like the Sphere, but when was the last time you saw Office or Windows significantly improved by that R&D spending?
In Microsoft's defense, perhaps we've tapped out the desktop software metaphor, and there's simply not much it can do there (beyond building SharePoint and the next tier of lock-in services to guard its cash cow product lines). Unfortunately, this "defense" is also my biggest critique of Microsoft: its future is so tied up in protecting its past that it's unlikely to ever unleash true innovations from the labs that could destabilize the desktop.
If you believe, as I do, that there's a bright future beyond the traditional desktop, it's hard to get excited about Microsoft's R&D spending, knowing that it's likely to lead to more of the same, with the occasional circus curiosity like Sphere.
Disclosure: My company, Alfresco, has a product that competes with SharePoint.