I would have preferred to respond in public earlier, but was muted at the time. Thank you for including me in this conversation.

You take issue with my tone, and with MacKay's. I take issue with using the words "very little land" and "tiny" to describe the space requirements of solar. But physics, not sentiments, will decide whether solar is viable. So I enthusiastically support running the numbers.

MacKay gives a peak solar energy of 1000 W per square meter, over 90 W per square foot. So we can hope that the 7 watts of today's panels will continue improving.

Seven watts per square foot gives 195 MW per square mile.

So 2.9 terawatts / 195 megawatts per square mile = 14,871 square miles.

How big is that? It depends on what you compare it to. I find it disingenuous to compare it to the continental United States. A test range is also simply a shape drawn on a map.

A solar installation would be an engineering megaproject; we should compare it to other engineering megaprojects. Small towns are mere dots on the Nevada map, the Hoover Dam is invisible. The floor space of the Pentagon is a fraction of a square mile. Fresh Kills landfill is four square miles.

No one thinks Elon Musk actually wants to build a solar facility the size of New Jersey in the desert. But he is definitely advocating when he characterizes solar as realistic for more than a small portion of our energy needs. If he wants to spend his own money on energy innovation, I salute him, but he is clearly willing to spend other people's money.