Applied to practical PV and wind power and batteries, this has some merit. But, the point is those lab bench designs HAVE led to structural improvements, engineering at scale. And they did in nuclear engineering too. All he was really doing is drawing a line through "simple, cheap and easy" language coming out of benchtop work.
It's like "... in mice" for medical advances. Taking the next step into humans demands rigour.
But is it the same at all? You design a solar panel and try one and make prototypes and then install thousands and it’s about the same and it scales. A nuclear reactor on the other hand…
A tonne of peskovite lab demos have failed to launch at scale in industrial deployment. It is really hard to make stable, high yield, high power PV. Small increments are all we're left with.
You're saying the risk:consequences equation differs. I don't disagree.
Modern silicon PV cells are cheap, non-toxic, stable, and their efficiency is more than half of the theoretical maximum. The big challenge is storage, PV generation already outperforms any other power source.
It's like "... in mice" for medical advances. Taking the next step into humans demands rigour.