> the scale that nuclear can achieve in 5-20 years

This is just fantasy. "Can" is a very long way away from "will". And almost all of the reactor projects currently under construction have time and cost overruns.

The best case for the UK in five years is that we manage to finish constructing one (1) reactor that is currently under construction.

Worse yet, it is the only one of eight plants announced by the government in 2010 to have actually begun construction (which started at the end of 2018). And 2010 wasn’t even the first government announcement about wanting a new reactor.

“Can” is a very long way from “will”. But “impossible” is even further. That’s my point.

Wind and PV grew by roughly 200-240 gigawatts in 2020. Total current worldwide civilian nuclear reactor electrical capacity is about 400 gigawatts, with another 60 GW under construction and about 100 GW planned.

Even if renewables stop their current exponential growth and only grow linearly, even assuming a mere 25% capacity factor for renewables and 100% for nuclear (neither of which is true), even assuming all currently planned and under construction reactors are completed and no existing reactors are turned off, renewables still exceed nuclear by about 2028. I’m expecting it to be a lot sooner than that because I expect the two (different) exponential trends in PV and wind to continue for at least another doubling period.

I’m still in favour of expanding nuclear power: it’s much safer than generally believed, it saves on storage requirements and intercontinental HVDC (we might get enough battery factories or extra copper/aluminium mines in time but I don’t want to count chickens before they’re hatched), and while it’s expensive it is (currently) still cheaper than batteries.

But I’m not expecting nuclear to be a big player, I just like diversity of supply for its own sake.