If you run the cost optimization with wind/PV/batteries/hydrogen at https://model.energy/ with 2030 cost projections (some of which have already been met) it comes out cheaper than nuclear at providing baseload (and hence at supplying anything, since baseload is nuclear's best case.)

The "running the numbers" you refer to very likely is a scam argument that requires batteries be used as the only storage mode. This can be much more expensive than properly engineered 100% renewable solutions, especially at higher latitude.

That website attaches Alaska to the lower 48 and doesn’t have independent grids for the lower 48 either. Wtf?

lol, after selecting a state it suggests a hydrogen electrolyzer and 22GWh of hydrogen storage. I hope you realize how dumb it is to think that would be achievable in 2030 (< 6 years from now).

If you don't like that, feel free to use the option for individual states, or the option where you draw an arbitrary polygonal region. Go wild!