Indeed - Sun stopped caring in the early 2000s. They were a great company once, and had great hardware, a great Unix, and great ideas about networked computing. But by the time they died, they had an ancient userspace with a nifty filesystem and kernel debugging tool, slow hardware, and didn't have the cash or skills to execute their big concept ideas.
I still find it hard to believe that a website with no income whose sole purpose is to allow you to throw virtual sheep at your friends can have more value than all the real engineering companies in America put together - at least for any sane definition of value!
You're simultaneously believing the hype about Facebook's valuation while discounting its revenue potential. With different assumptions I suppose one could arrive at a different conclusion.
It's difficult to see a revenue potential that justifies a $50Bn valuation. There are 2Bn internet users at the most generous estimate - for half of those $1 is a large amount of money.
It's difficult to see how advertisers are going to extract enough value from the other half to make paying $50 to Facebook for each of them likely. When even a the most expensive superbowl commercial only costs 3c/viewer.
Unsentimental redeployment of resources to newer, more promising ventures is one of Silicon Valley's competitive strengths.