> Comparing a cost that's been validated across many units over decades with the "should cost" for a design that hasn't even had gone critical in a prototype is almost meaningless.

The poster is likely talking about new fission concepts like TerraPower [0] and other Gen IV types. See the Generation IV fast reactors section of [1].

[0] https://www.terrapower.com/about/

[1] https://www.world-nuclear.org/information-library/current-an...

I wasn't but it applies to that as well.