Construction, Mining/fuel processing, Operation, Decommissioning, Fuel Storage / reprocessing, Disaster Insurance. Sure, we can hand wave R&D as a sunk cost, but many of these costs only show up after the fact.
All of that stuff is debatable of course but given the massive environmental damage done mining Lithium I doubt there is much in it from the point of view of that one way or the other.
As for fuel storage/reprocessing with modern reactor designs the waste products are much smaller and we do have sensible sane ways to deal with them.
I'd absolutely love it if renewables dominated the percentage for generation but I think the future is more likely to be nuclear for baseload and renewables for the rest barring some huge breakthroughs in battery technology.
Pumped hydro is still the most efficient way of storing huge amounts of potential energy but it's simply not a possibility for most places (and is hardly kind to the environment either since it often requires damaging lakes and rivers).
Nuclear isn't perfect but it looks like the least worst option to me if implemented properly.
You can send electricity 1,000 miles with minimal losses. So there are some areas outside of reasonable pumped hydro but far fewer than you might think.
Construction, Mining/fuel processing, Operation, Decommissioning, Fuel Storage / reprocessing, Disaster Insurance. Sure, we can hand wave R&D as a sunk cost, but many of these costs only show up after the fact.