I think what he's asking, and if not, I'm asking out of my own ignorance on the topic. If solar isn't cost effective on small scale installations (home use), what is it that makes it cost effective on large scale installations? Is the return not linear?
I think he's confusing the technologies here...
Let's go through this briefly:
1) We already have a nuclear fission based economy. 20% of American electrical consumption comes from nuclear power.
Many European nations have much much higher usages of nuclear power. France, for instance, in 2008, drew 76% of it's electrical consumption from nuclear power, with only 12% from fossil fuels.
Looking at France, specifically; here's a snapshot of French electric generation (different than consumption, fyi):
Here's energy consumption (this includes cars/trucks/etc):
As you can see, the problem here is not fossil fuels being more efficient over nuclear power (they're not); but instead, the reliance on the internal combustion engine for locomotion. Move consumer driven automobiles to primarily electrically driven engines and you've removed the vast majority of fossil fuel energy consumption.
With respect to questions of cost, while throughout Europe, the cost per kwh for electricity floats ~$0.30, in France it's $0.169; comparable to the United States at $0.125.
The argument here is not that these technologies be emergent from the free market (which is what would take 40-50 years), but instead, that the government should be subsidizing the expenditures, just as it has for decades with fossil fuels; thus shifting our economy to clean energy.
2) Solar thermal plants already exist; here's one operating right now:
That's Ivanpah, a 400 MW solar-thermal power facility. There are no photovoltaics in use here. The plant's operation, in the most simplistic explanation, uses mirrors on the ground to focus light back at the central tower, at which, boilers generate steam to turn turbines to generate electricity.
This plant is new, but using decades old technology.
These kinds of ideas have been around since the 1950s, but we've moved away from innovation like this in favor of protecting corporate profits.
3) Space-based solar power is the
end-game. Such a system would use photovoltaics in space, beaming down microwave energy to the Earth's surface. It would require no new technology, but instead, massive infrastructure to be put into place to maintain these arrays. However, SBSP could power the entire Earth well into the foreseeable future.
These kinds of technologies already exist, yet, like our space-program, have been mothballed because of government ineptitude. Which is the entire point.
We don't do these things because they're impossible or infeasible; we don't do them because we choose not to.