I expect we'll overbuild renewables to handle the high demand season and then dump excess renewables in the low demand season. I've seen this result in a couple cost optimization models like [0]. Optimistically, we might see demand react to the seasonal price differences and readjust via e.g. weatherization, or running industrial load seasonally.

But yeah, seasonal storage makes my head hurt. Getting 1 cycle a year out of technology built to deliver 2,000 or more cycles seems like a huge waste. The idea of financing a project to pay back over 2000 years is bizarre. Of course the hardware would fail for other reasons over that period. Hell, the concrete it's resting on would probably crumble in 100.

[0] https://kleinmanenergy.upenn.edu/research/publications/feasi...