The National Research Council (via cnet) brings some much needed reality to the debate about "green" energy:

...renewable energy sources--wind, solar, geothermal, wave, tidal, and biomass--could supply 10 percent of U.S. electricity supply in 2020 with existing technology.

Getting to 20 percent of U.S. electricity by 2035 is possible with sustained policies and investment.

Another important, but largely ignored, aspect of the problem is a bad fit between today's grid and renewable sources. Over the last 100+ years, the grid evolved to accommodate industrial-scale, reliable, on-demand generators. Now, we are trying to augment it with relatively small, inherently unreliable, when-the-sun-shines-and-the-wind-blows sources. Bad thinking. It commits a very common problem-solving mistake when people attempt to "graft" a piece of advanced technology onto an old infrastructure. Doesn't work. The new energy sources require new infrastructure and new distribution logic. I would argue that energy storage and supply/demand localization are much more important here than solar panel and wind turbine efficiency. It's the system, stupid!