Yes, but in the last 30 years, we've went from prohibitively expensive 4.5% efficency cells in 1954 and $1500/watt in 1955 to a test 500kW installation in 1977 to $9.00 in 2007 to, in an ideal situation, 8.5 cents per KWH today.
We could start building a 100% solar/nuclear solution in the US today. The technology is there, and it would only be the construction time. Our energy costs would go up, but not significantly. However, if the technology keeps on this curve, unsubsidized solar will be cheaper than coal soon.
We aren't doing this today because the power companies know that the curve is coming soon. Why invest prematurely, outside basic research projects?
Solar is already in plenty of niches - from highway signage to parking kiosks. Every niche helps increase production, which reduces price per unit.
This is why I support subsidies - unlike for oil or nuclear, they're actually working to get costs down.