Ask any competent security researcher which is preferred: open critical review or security by obscurity. They will all tell you the former is best. The notion that proprietary software is somehow more secure due to its opaque nature is nothing more than a nod for security by obscurity, and is wrong-headed.
Would you rather have many good guys and a few bad guys pouring over open source software looking for flaws (with the good guys pointing out and fixing those flaws), or just have the few bad guys reverse engineering the flaws in the closed software and of course not telling the vendor? How often do you hear about day zero exploits on proprietary software versus open source?
Steve Ballmer certainly is an intelligent, forward-looking, successful business executive and we need to listen to his visionary thoughts. On the other hand, it is sometimes hard to tell if Steve's statements are biased towards where Microsoft just happens to be in the market race compared to where Steve would like them to be.
Clearly, any major player cannot ignore the pace of mobile technology demand so Google certainly can't be considered "wrong" in their direction. It is certainly conceivable that Microsoft, a provider of $60B of desktop software and server databases, internally, considers themselves to be late in moving a huge ship in the best direction that will sustain them for another 20 years.
IMO, based on a recent MS meeting I attended, MS does not have a clear message for business executives. A case in point is Sharepoint. While many companies are interested in buying the product, it is not yet successfully positioned nor branded to address strategic business issues that many companies will be facing over the next 5 years around communications, collaboration and content management. Let's remember that MS Windows beat out IBM's more technically capable OS/2 due to innovative marketing, not product functions and features. MS is subject to a similar fate as IBM unless it figures out how to get the game back on their playing field and exercise their once unparalleled marketing prowess.
Computing and computers are becoming so ubiquitous that the line between personal and work PC is all but faded away. We do need to have control over the assets, but the assets are electronic documents. If we control the access and flow of t he documents, why do you care if the PC is yours or your company's?
An IT project is no different from a civil engineering project when you're talking about complexity: The bigger the building the bigger the risk, but it does not mean we are going to build only small buildings.