Earlier quoted context omitted.
Don't forget Germany. If you look at the amount of PV built in Germany early this century and make some admittedly strong assumptions about learning curve, one could argue the Energiewende, then usually called failure, singlehandedly accelerated PV development by decades. I don't recall Germany ever credited on that.
I remember some old tidbit about the American westward expansion, most railroad projects failed and went bankrupt and were sold for pennies on the dollar to the ultimate owners. Something sad about that, really.
The lesson, which we learned in the dot-com era and will likely learn again in the AI era, is that the benefits of step-change new infrastructure technology do not accrue in the long run to the infrastructure builders—the technology only creates the step-change if it finds its way to being a commodity!—but diffuses throughout the new, ultimately much larger, more productive economy as a whole.