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 g

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.