The problem with quantum computing is that it doesn't work well for everything, but what it works at (Shor's algorithm aka factoring gigantoid numbers, or creating secure OTP streams), it excels beyond belief. I'm betting that the near future of tech will be more or less how Google as lived: Lots of cheap systems put together to make one giant entity. Right now we're seeing some SMP, SLI, distributed systems (OpenMOSIX, LusterFS) making headlines, but they're still a bit aways from being more commonly used.