Considering the ~US$187 billion dollar cleanup effort after the most recent failure, just let them try can be an extremely expensive proposition. Operate for 50 years then say, sorry cleanup is your problem now suckers is similarly an issue.
https://en.m.wikipedia.org/wiki/Fukushima_disaster_cleanupThat’s the catch 22, you need to not just be extremely safe but able to convince others outside your organization that it’s going to work. Having been burned twice by failure modes outside the original context it becomes even more difficult to innovate. Novel designs have novel problems and unknown unkowns can be a huge deal.
Can you get away with fewer guards? Well probably, but the minimum is not clear. How about thinner walls, again probably but can you convince others it’s a good tradeoff? Now extend that to everything.
