That said, I think he overstates the case here:
The situation is similar for solar thermal technologies; they have had major R&D expenditures for decades and they are improving slowly. But they can never be as cheap as coal-fired electric generation because the energy density of the sun’s rays are not nearly at the level of fossil-fuels like coal, so you necessarily need more physical equipment to collect the energy, and turn it into electricity. Also, the lower temperatures that result from collecting the sun’s rays compared to burning fossil fuels inherently limits the efficiency of solar generation, but more importantly, it increases its costs relative to fossil generation.
It's very difficult to predict, based on physics, what kinds of technologies will be cost-effective and which won't. Physics would suggest, for example, that long-haul trucks could never compete with trains for hauling heavy freight (in fact, they dominate the market in the US). Rosen is, under the gloss of a scientific argument, reasoning ex post facto from the actual relative cost of these technologies today.
Suppose we instead lived in a world in which wind, photovoltaic, and solar thermal sources provide most of our energy, and some clever reformer was proposing coal as a solution to the intermittancy problem. How might Rosen explain the prospects that coal would overtake solar and wind?
ALTER-REVKIN: Coal is a promising emerging technology, easily scalable, with theoretical efficiencies twice what we can achieve with solar, but will it ever compete on price?
ALTER-ROSEN: Coal may be a valuable minor player, but it will never be as cheap as solar. There are too many costly inputs and costly side effects. Imagine, you need a research team to locate the coal, you have to purchase the rights to the coal deposit, then you need an entire operation, separate and independent from power generation, to get the stuff out of the ground. That's trucks, it's heavy machinery, burning fuel and writing paychecks to the operators. Then you need to haul it to the power plant -- more trucks, more heavy machinery. Finally you burn the stuff, and it produces coal ash, which is toxic. You have to store that safely for hundreds of years -- it's not like you're going to dump it behind a rickety wooden dam somewhere and walk away!
ALTER-REVKIN: Wow, that's a lot of costs.
ALTER-ROSEN: And we're not done yet. Researchers estimate that if this technology were widely adopted, millions of people would die each year from atmospheric pollution. So the companies would be paying from that, as well.
Bottom line, it's too complicated to find it, extract it, transport it, store the wastes and cope with the consequences of the pollution for coal to every compete with a no-fuel, no-pollution source like wind or solar.