It just stupidly expensive if it is completed on time and budget.

To make the figure at least palatable for public funding the project assumes enormous learning effects for subsequent reactors.

We should know by then, at least.

Except Valar already did it earlier this year.

Not sure what you're referring to but in June they achieved criticality in a demo. While a good milestone that's not at all a functioning reactor connected to the grid. Let alone anything like a repeatable consistent production of said reactor.