Let me make some things clear, I am all for using truly renewable energy. The large (450 kW, I believe) wind turbines sold by Siemans & GE are something I developed for the (US) Department of Energy (DoE) back in the 1980's. (The story is complicated, but I saved the U.S. Windpower project from cancellation, put together a consortium of small companies, and built and delivered 987 units to DoE for less than (McDonnell-)Douglas Energy Systems' proposed tooling budget.) I also have designed, built, and installed wind turbine systems in the arctic and North Dakota. I have designed and built solar ovens for 3rd world usage, solar forges (concept demonstration & testing), and solar water heaters that have provided 65% of year-round 180°F water in Albany, NY. I have also participated in 3 PowerSat design & development projects. The potential is huge.
The problem is that the costs can also be huge if things are improperly designed or implemented. As noted previously, one of the companies for which I do design/engineering and development work is a (US, Swedish, & UK) government subsidized producer of amorphous silicon PV systems. They run about 12% efficient on average meaning that, at full tropical solar flux (i.e. 1 kW/m²) input, they produce about 120 W/m² of output. They cost close to $1500/m² to produce and have a lifespan currently running in the 10-year range. They also require (shall we say) not inexpensive power conditioning before their output can be used in standard AC applications. On a bright sunny summer's day here in Washington State, a good output from them will run in the 50 W/m² range. If you let them sit uncleaned here in the Puget Sound region of Washington State for six months, their efficiency drops from 12% into the 4% range.
There is a technology under development in Japan that, as of about a year-and-a-half ago when I sat through a meeting on their progress, is reputed to reach somewhat above 40% efficient in PV conversion. The manufactured cost for this unit was (at the time of the meeting) running about $12,000/m² with a lifespan reported to be 3 years. The costs will eventually be reduced and lifespans increased, but that was the state of the art the last I knew.
In terms of solar forge (i.e. power tower) technology, the problems still to be overcome are far from trivial. There was a really kewl approach being tried in the early-/mid-70's that still appeals to me -- there was a variation on a sun-flower developed that was reputed to be more than 80% efficient in terms of reflecting infrared (IR) wavelengths. The flower itself tracked the sun based on UV wavelengths, so a UV hologram properly positioned could keep the IR focus of the flowers on the thermal collector. The flowers themselves would close-up at night and self-clean their reflective surfaces. I have no idea what happened to that program. It was based in one of the Persian Gulf states and got "moderate" (i.e. you had to dig for it) coverage for several months and then disappeared from sight.
The world's energy consumption in 2008 averaged (about) 10 TW-hr/hr (10,000,000,000,000 W-hr/hr). That implies that peak energy consumption ran somewhere in the 30 TW-hr/hr range. If we assume that we can ultimately develop a system that is 50% efficient (the best you can do under the Laws of Thermodynamics) and that we are working with an input of 1000 W/m² that is available for an average of 12 hours/day. That implies that, to meet our total energy needs, we would dedicate 2*2*30,000,000,000,000/1000 = 120,000,000,000 m² (or 120,000 km²) to that generation. That is a square (roughly) 350 km on a side -- and then you have to distribute this power.
To use a rather obscure analogy, rail transport is (approximately) 50X more fuel efficient than truck transport and (approximately) 200X more fuel efficient than air transport. Here in the U.S. about 20% of our railbed system is out of service due to lack of maintenance. Americans spend more than 300X the tax-dollars subsidizing truck transport and nearly 10,000X the tax-dollars subsidizing airlines as we spend subsidizing rail service. Add onto this burden the (well-publicized) hucksters (such as Solyndra) who have "fed" at the tax-dollar trough of alternative energy subsidies, and you start to see the real nature of the problem.