Something I've mentioned before is that uranium-based nuclear currently isn't scalable due to the limited amount of extractable uranium that's estimated to exist. 230 years worth, but that includes about 80 years
known reserves, plus the 2/3rds of unknown reserves they predict that we will find eventually. If nuclear is scaled up
now to solve the CO2 problem, then you're looking at the 80 years worth, and if you quadruple current nuclear capacity, you're down to 20 years worth of viable known reserves, i.e. less than the lifecycle of the plants. So they'll take too long to build, and even if you fix that problem by somehow safely fast-tracking massive numbers of reactors, they burn through the viable fuel too quickly. And even if you solve that problem you have the problem that you're now generating nuclear waste way faster than we know what to do with all the nuclear waste. "reprocessing" is currently science fiction. We can barely "reprocess" what we already generate, and it's a big problem. And not for lack of trying. Anyone who works out how it's viable to turn nuclear waste back into fuel is going to be a trillionaire.
The point is, if you
can't currently scale up nuclear by over a factor of four without more or less running out of fuel instantly (lasting a whole two-Trumps worth of time. New unit of time. A Trump = 8 years), then it's not a viable solution to replace coal and natural gas.
We can scale it up only with future tech, and by that time, we won't need it.
https://www.scientificamerican.com/article/how-long-will-global-uranium-deposits-last/EDIT: I'll also mention that we artificially boost off-peak usage with variable pricing, to allow more usage of large power stations at night. "Base load" power is thus artificially inflated to smooth things out. If things switch to solar being more dominant you could merely reverse the pricing incentives. And while it's not guaranteed to be sunny where you are, long-distance power transport is surprisingly efficient, only losing about 1% per 160 km. So it might not be that sunny in one place, but the grid evens that out over many places, and 1% per 160 km to transport power is more efficient than storing in a battery. So you can have reverse-off-peak pricing to shift usages towards the sunny times, along with sharing power among the grid across cities, to even out fluctuations in sunniness. Since "base-load" would be a lot less in this scenario than currently, existing base-load sources such as nuclear can continue to service that without needing more capacity.