Ditto for air flight: from the dawn of aviation to about 1960 or 1970, you see steadily increasing speeds. Today, most planes still fly at about ~500 MPH (I think), just like they used to.
Although there are still lots of micro-innovations going on with cars, rail, and planes, they've basically hit a maximum airspeed.
I suspect something similar might be going on with the current paradigm of desktop operating systems. Since around ~2005 (when Windows XP's service packs finally caught up to virus / worm writers and OS X Tiger brought stability and full text search), nothing in operating systems has made a tangible difference in my life.
Compared to the jump from DOS to Windows 95, or from 95/98 to XP, or the late versions of Mac OS to the early versions of OS X... the current changes in operating systems just aren't that great. Which isn't to say "consumer OSes are dead"—but they've probably hit a spot where the major problems have been solved. The innovations are smaller than universal protected memory, fast, full-indexed search, and the like.
The key words above are "for this paradigm." Mobile represents the next obvious computing paradigm, and that might be a good thing. For better or worse, Microsoft has "won" the desktop and will probably never lose it. The company appears to mostly be competing with itself these days (I say, writing from an iMac).
Multitouch in desktop OSes has been tacked on as a layer over the traditional cursor-oriented APIs, which is why it feels that way when you use it.
Not necessarily. It could also mean one cursor that operates in four spacial dimensions.
It's interesting to see this stated now, right as Google is moving forward with automated cars. Once all highway traffic is automated we can push to much higher speeds safely (and drive much closer to each other).