One of the problems in the US is that there are research reactors and production reactors, nothing in between. Research reactors are to small to prove out many concepts.

So startups have to essentially go from nothing to first commercial reactor in one step, without iteration.