20 years ago, solar panels and electric cars also had "scale" problems. Those turned out not to be problems.

No, even forty years ago it was clear that you could power the entire USA from a negligible sacrifice of southwestern territory to solar power. Your analogy doesn’t work.

You could, in a world of spherical cows. We could just as easily say synthetic fuels could power the entire USA if we just scaled it up.

My point is you can’t make that argument. It’s obviously false. You would need millions of the facility in the press release to even make a dent in global liquid fuel supply. You would cover the earth with windmills before you finished the job.

On the contrary, if you looked at solar panels in the 1990s, you would think that there would be flying cars before you could scale manufacturing up to fill a good chunk of the southwest US. Only later did people actually make silicon solar panels that could be cheap enough, and it took longer to make them with good enough yield to actually deploy them in large quantity.

This technology doesn't seem to have many known scaling limits yet, just processes which haven't been well-researched, so there's no need to dismiss it early.

I mean, technology tends to get smaller, cheaper and more efficient over time.

It's like objecting to the first computers because they took up a giant room, were heavy and could only do 3rd grade math. They'll never go anywhere.

Conversion of atmospheric CO2 to liquid fuels isn’t ever going to get more efficient. It’s limited by physics. And they claimed future capacity of that plant already bakes in substantially optimistic assumptions about benefits of scale. Its current capacity is 100x smaller.

If the technology can be miniaturized and applied directly to a car, it would generate its own fuel. Combined with an electromotor/batteries as a hybrid engine, this could be an extremely efficient way to drive.