Second, Solyndra failed because it couldn't compete. The costs of solar panels are dropping, which is a good thing, but not for Solyndra, whose panels were pricey.

Shugar said the costs of solar panels have dropped from about $6 a watt in the mid-1980s to about $1.25 per watt today. "The cost of solar has dropped ferociously," he said.

Costs continue to fall. As the Lawrence Berkeley National Laboratory recently reported, the price of solar installation dropped 17 percent from 2009 to 2010, and another 11 percent in the first half of 2011.

Third, Solyndra and the rest of the solar industry remain entirely dependent on subsidies. Solyndra got the now-famous loan guarantee. Its customers enjoyed the 30 percent investment tax credit. What's more, because more than half the states in the U.S. have renewable portfolio standards, requiring utilities to include renewable energy in their mix, solar often doesn't have to compete with coal or natural gas but with other forms of renewable energy, principally wind; the utilities can pass the higher costs of solar onto their ratepayers. [See my 2010 blog post, The Hidden Costs of Solar Power, written after I visited Solyndra.

Republicans are threatening to repeal the 30 percent tax credit. "That would be a terrible tragedy," Harris said. "It would put at risk the momentum this industry has created."

Interestingly, Shugar and Harris both said they would be willing to give up or sharply reduce the 30 percent tax credit as the costs of solar continue to fall, so long as subsidies for competing technologies (nuclear power, natural gas and coal) were eliminated, too.

"Solar is going to reach the point where it can be competitive without the investment tax credit in five years or so," Harris said. Shugar agreed, saying: "We can certainly talk about removing all subsidies. Over the longer term, definitely, solar can stand on its own."

In the meantime, the single best argument for clean energy subsidies is not "green competes" or economic competitiveness with China but the fact that fossil fuels enjoy an unfair advantage over solar and wind: They don't have to pay the costs of emitting carbon pollutants, which are leading to catastrophic climate change. Taxing or regulating CO2 emissions would be better, simpler and more efficient subsidizing clean energy, but that's not going to happen anytime soon.

Unfortunately, the incentive tax credit is an imperfect and wasteful subsidy because it's cost-based. If I spend $30,000 to put solar panels on my house, which is shaded by big trees in Bethesda, MD, and my contractor overcharges me, I get the same tax break as an Arizona homeowner who spends $30,000 wisely and generates more environmental benefit. That's wrong. There should be a way to subsidize clean energy based on its performance, as the NRDC's David Goldstein argues here.

More important, though, I'm persuaded by the Solyndra story that there's no compelling reason for the government to provide loans, grants, tax breaks or favors to individual companies, no matter how worthy they may appear to be.

Picking out promising technologies and management teams is a job best left to venture capitalists, whose success rate, by the way, is not especially high.

For a lot of reasons -- the temptation to play political favorites, the risk of outright corruption, the fact that worthy companies can raise money in private markets -- it's not a job best done by experts at the U.S. Department of Energy.

Particularly because they're betting with other people's money.