There are three long-term solutions: one is to greatly improve electric transmission so inconstancy is averaged over a larger area (ultimately Buckminster Fuller's idea of a world grid). The second is greatly improved (and cheper) energy storage - enabling electric vehicles provides some other incentives for that too, and may actually be part of the solution.

The third long term solution is to go where the sun does shine constantly - off-planet. Space solar power in various forms has been seriously proposed for almost 30 years now, with some minor demonstrations of transmission feasibility etc, but not much real research effort. The Japanese are spending a few million dollars on it right now though - there's a suborbital experiment on zero-g robotic construction coming up in the next few weeks that should be interesting.

So yes we have some real challenges, but so does nuclear power. Let's spend the money on real R&D now, and let them all compete on as fair a basis as we can make it. It's my belief that one of the primary energy technologies (probably not wind, and not once-through nuclear fission) will become the dominant energy source of the future, with a majority of the market thanks to cost effectiveness. But none of them are there yet; we need a lot more R&D.

http://www.geoexchange.org/

The Comparing Systems page gives some useful numbers.