You can also see it conceivably helping with a start up idea that pushes the boundaries, something that requires you to do years of research to pull off well is probably going to be a bigger gap in the start up market that something that just does something in a crowded market slightly better and more efficiently.
With a CS PhD, you have a lot of options: academia, industrial research (Microsoft Research, IBM Research, and so on), working for or founding a startup, etc. Happily, many CS graduate schools and professors explicitly recognize that many grad students don't intend to pursue a career in academia, which I think is less the case in other academic fields.