Yes, I am aware of subsidizing R&D but that has diminishing returns too, and given the amount already poured in around the world and the rather dismal returns, I'm underwhelmed by the proposition that pouring even more in will turn things around. You can always claim that if you just keep pouring the money in it'll all turn around; it's a null argument when it comes down to it. (We'd almost certainly be better off pouring equal funds into getting nuclear going instead.)

Extractive energy production where you can dump negative externalities onto the public or hide your subsidy in a part of the budget that is not directly traceable to you (e.g. Marines in Iraq and destroyers in the straits of Hormuz) is surely one of the purest wealth producers, but if forced to actually compete on its own the equations would look a bit different...

Wind and solar both generally barely break even or barely above if you take a full accounting of their energy inputs and costs, biofuels are often a net loss (depends on the crop, but I think the balance of the argument has corn ethanol as a net loss, cane sugar seems to be a net gain, but...), and the problem is they're competing with things that easily get tens of times of returns on energy expended with the fossil fuels and nuclear power.

One of the things you rarely see correctly computed is what it would truly take to power our entire society with renewable energy, including the sudden new energy expenditures necessary to keep our purely-renewable infrastructure maintained with replacement gear. As the net energy benefit of the average piece of gear approaches 1x, the necessary expenditures approach infinity. Replacing 10-25x sources with 1.5-3x sources requires yet again far more resources than the naive multiplications and divisions would imply, if you don't make the mistake of assuming free infrastructure that never decays, or one-time-cost infrastructure that never decays.

(Incidentally, this is why cheap solar, in the sense of truly cheaper without government subsidy solar, is exciting. A solar panel that can make back 5-10x the expenditure to make and install it, and isn't a massive expenditure of metal and glass and silicon is a big deal, it makes things practical that weren't before. Or a solar installation consisting of lots of cheap reflectors concentrating the energy on a centralized station. I still think we might be able to go both net positive and practical on solar. Wind I'm less optimistic about, it's difficult to see what we can cut out of our wind generators and still have wind generators the way we can cut down on the mass/energy footprint of a solar installation with clever engineering. In the limiting case, a reflector is a sheet of foil and an amortized central station; a wind generator is an entire wind generator.)

If I were John Boehner (leader of the US opposition), I'd take money from corn and oil subsidies and use it to fund more R&D, but I'm not (or at least I won't admit to being him in this public forum).