I think your key assumption here is wrong: uranium isn't fundamentally expensive. It's expensive because there isn't enough current demand for it to bring more uranium mines and enrichment facilities online. Uranium used to be cheaper (adjusted for inflation) per gram than it is now, because there used to be more of those facilities online than there are now. With increased demand, it would be cheaper again.
> Uranium used to be cheaper (adjusted for inflation) per gram than it is now, because there used to be more of those facilities online than there are now.
This is both false and entirely irrelevant. The costs are driven by capital (and require supply chains that don't exist) not fuel. Minerals get more expensive to extract after you extract the easy stuff. Building out 100s of GW of new nuclear would require extracting the stuff that costs several times more than present -- to the point where fuel costs would be equal to the LCOE of solar.
What is relevant is the entirety of world reserves are not enough to provide even US electricity + transport energy in PWRs. Loading 800GW of reactors uses almost all of it. Reprocessed MOX and what's left might buy you 20 years of operation. Mines take quite some time to come online so the pace of production of new nuclear would be small compared to even the torpid rate of US renewable production.
The PWR industry is nowhere near the scale of renewables, and it's impossible for it to get there.
If you want to blow a trillion more dollars on trying to make it happen, put it into liquid sodium FNR research. At least that kinda-sorta works. You'll be quite disappointed when you finally get a design that is safe and scalable and see the price tag though. And even if you do go all in it will take decades to breed enough fissile material to make a dent.
