Thanks for that. The main criticism is that "it looks like a curve fit". Indeed it does; the paper talks at length about this very thing. (Mostly in the 400 pag

Thanks for that. The main criticism is that "it looks like a curve fit". Indeed it does; the paper talks at length about this very thing. (Mostly in the 400 page supplementary paper S1.) But the thing is, the curves (Moore's and Wright's) apply to a lot of technologies undergoing rapid expansion. That's actually the science contribution of the paper: setting out a better way to make policy-relevant forecasts, and hopefully ending the abysmal track record of forecasts that is documented in the paper. (I can see ways to get at least 67% cost reductions in both wind and PV from where we are now, and I'm just an interested layman. So in at least the near term we will likely keep following the curves. Yes, actual work has to be done, but there's no voodoo. Just like with microprocessors.)