> Grid scale lithium batteries have an effective lifecycle of 15 years. In this potential future, global lithium reserves would actually start getting choked up before the 2050 goal.

I think the long-term solutions here are not grid-scale lithium batteries, but pumped hydro, flow batteries, or compressed air. Lithium batteries have just gotten a bit ahead on the technological growth curve because of the recent boom in production from phones and EVs, but liquid flow batteries can be made using common elements, and are likely to be cost-effective once the tech gets worked out better.

So: I don't think we can say "lithium energy storage is unfeasible large-scale and long-term" and thus conclude that nuclear is inevitable, unless we also look at all the other storage alternatives.

We already have an electric grid we don’t need to build a new one from scratch just replace infrastructure that gets to old and add more for whatever extra demand shows up.

Obviously other energy sources are going to exist and non solar power will be produced, but nuclear is getting fucked in a solar + battery heavy future. Nuclear already needs massive subsidies and those subsidies will need to get increasingly large to keep existing nuclear around let alone convince companies to build more.

> That's assuming constant production of batteries at the current scale and production.

That's a terribly pessimistic assumption when production has been scaling exponentially, and cost per kWh dropping exponentially.