Ironic really. At it's inception, microsoft got in on the paradigm transition from mainframe to desktop computing, albeit not through any real innovation or offering of value. IBM and mainframes are still around and arguably profitable, but not relevant in terms of what drives current trends of innovation - at that point new and interesting directions started coming from the desktop computing world. Think of every thing you do on a desktop computer now, and it's likely something that didn't arise from the IBM/Mainframe computing world. Now things like google glass, new distribution models like hadoop, hyperscale computing, and "cloud" computing (yes I held my nose when typing the "C" word) are arising in a sense from the budding influence of the mobile computing world.
So now the paradigm is shifting to mobile, and MS has missed the boat in almost every definable way. But like IBM, they'll stay around and still be profitable, but they just won't be relevant in terms of new directions in the information landscape. The apple didn't fall far from the tree, no pun intended.