The proposition here is to throw a lot of time and money into a design that might not even work satisfactorily, and might not ever make any money even if it does.

The first is for practical reasons. I'm sure the theory works. But what if the practice is something like a pipe leaking, a liter of highly material spilling on the floor, and it being too contaminated to touch for a year? Yeah, that won't kill anyone, but won't be good business either. There's lots of possibilities for it working perfectly as promised, but just having various issues that make it non-viable anyway.

And I have a very hard time imagining a design that could compete with renewables. Because no matter how you slice it, nuclear has complex requirements that say, solar can just completely ignore. This is especially so when you start talking about not yet production designs, so maybe the first one will be built in 15 years from now.