Comments by "DynamicWorlds" (@dynamicworlds1) on "Asianometry" channel.

Temperature isn't that big a deal, actually, at least not for grid storage. Ambri's losses were entirely manageable, and you're dealing with a square/cube problem. The bigger you make them and closer you pack them together, the less of an issue you have as efficiency increases with scale. Keep in mind also that all power distribution and storage has inefficiencies and you also save massively on reprocessing vs the maintenance and replacement vs other energy storage systems (like lithium ion batteries that don't handle deep cycling well and have short lifetimes with massive costs in extraction/recycling and assembly and all that takes not only money but energy (and some systems require active cooling because operation increases temperature which decreases efficiency) I don't know GE's numbers specifically, but considering you can square-cube your way out of heat-loss inefficiency problems for grid storage, I'm not concerned. It's only really a limit on practical minimum size for the technology, not the viability of the technology. It's not that often the square cube law works in our favor when making stuff so enjoy it when it does.
2