Or we could move our scientific computing workloads to places where that renewable energy is, as transmitting the results of scientific computation is much chea

Or we could move our scientific computing workloads to places where that renewable energy is, as transmitting the results of scientific computation is much cheaper than moving the energy. I'm not saying that Proof of Work algorithms have no value at all, but I believe they have very little value to humanity compared to most other uses for that energy, and increase the costs of energy in many cases for most other applications. Currently Proof of Work spends energy to create a betting market for rich people. Maybe in the future there might be useful application of it, but we've yet to get any further than theorising and small scale demonstrations.