The US Navy's demand for reactors is insensitive to cost. The tactical considerations of needing to refuel only once per decade (and in the case of subs, being quiet) outweighs all other factors. What works in the context of the navy doesn't really have a bearing in the context of commercial power generation.
And you're saying there's no means by which more economical versions of these could be built? With virtually all technologies, even complex, difficult ones, commercialization does create cost-cutting innovations. A modular, compact version of USN carrier/sub reactors without the worries about noise or extremely long refuel cycles seems feasible if the fundamental technology is already there and waiting to be (I hate this tech phrase so much but here it goes)... "disrupted"....
But the problem is that we know how to make reactors more economical: you make them larger. Maybe some smart people will find a way to bend the curve; I wish them luck. However, be careful in assuming that the aforementioned "cost-cutting innovations" via commercialization represent a free lunch. The sad fact is, most of the cost-savings from commercialization come from reducing the quality of the product, which is the one thing we don't want to compromise on when it comes to nuclear reactors.
