The problem isn't the perception of government backed businesses being 'wrong' for profiteering (or competing with industry?). Most people would be happy if the government made successful organizations that had a real impact (NASA in the 1960s). But successful large-scale government projects are rare today. Any gov tech gains in the past couple of decades were usually a minor side-effect of other massive investments (see: wars, intelligence gathering).
Modern nation states are simply not known for being good speculative investors nor good at running projects efficiently or cost effectively. Even when it's via private industry collaboration. But they are good at being politicians (ie, creating tax policy, managing regulations, etc). So my view is that they should focus on that and let industry be good at building technology.
I don't think the problem is government, the problem is funding - companies with limited funding would do an equally bad job of tech as a government department
Also bear in mind that assessing the 'cost of solar' is a toy problem as long as solar can't supply baseload power requirements.
The cost of solar will continue to crater; it is not a matter of "if" we will run entirely on solar, but only "when" at this point.
Also, Hawaii gets its electricity from oil and gas shipped in by tanker and residents pay $.30/kWh; the utility there managed to negotiate $.15/kWh from SolarCity, which is almost certainly below their cost of production right now; they couldn't meet the utility's original tender and reapplied several times. I'm guessing that SolarCity is betting that over the life of their 20 year (!!!) contract they can bring their costs down and eventually turn a profit. I wonder about that bet, though, given that SolarCity isn't in great financial shape.
Anyways, I'm very much in favor of subsidized and incentivized solar power; what grates on me are people who assume that there's a great deal of unrealized value in clean technologies ("new industrial revolution!"). Converting to clean technologies will destroy value because oil and gas, especially when untaxed, are cheaper and easier to exploit. But of course we must convert because, well, 400ppm.
Cumulative global installed solar photovoltaic capacity will surpass 310 GW this year, compared with just 40 GW at the end of 2010. Polysilicon module prices continue to plummet; even if storage costs have not caught up, we can continue to drive out fossil fuel generation with more solar and wind, not to mention shifting loads to renewable generation hours whenever possible.
Having nuclear power plants running all night when no-one needs the power is sub-optimal. That's why many nations built hydro storage and offered financial inducements for people to use power at night. Because the numbers wouldn't work for nuclear if it wasn't used at max capacity the whole time, producing a static amount, yet demand rises and falls both daily and with the seasons.
Doing the same thing, decades later, for solar is hardly rocket science, the only difference is the power correlates very well with air-con loads on both a short and long term scale which makes it a no-brainer for many locations around the world.