> I mean, we have deliberately initiated containment failures for military purposes, so we know what they can look like.

This is an empirical argument.

Is the upper bound of damage higher for nuclear reactors or fossil fuels?