Even though they do fail badly (when managed badly), they still kill fewer people per installed GW and produced GW-h compared to coal, gas and even solar.
When managed well (i.e. Adm. Rickover's US Navy) nuclear power produces zero health-related incidents over some 14,000 reactor-years. When managed super-badly (i.e. Soviet Navy) you get multiple deadly incidents per facility per decade.
I would say that calling this a "hydroelectric catastrophe" is misleading. The main motivation for building this dam was apparently to control flooding.
That first one might be of no use in a dataset for the US, UK etc. to make such decisions. It is a ridiculous outlier. 20-200k people die and there is a cover up.
Whether the government tries to cover up or not is of little consequence, the people are already dead.
Multiple 100+ fatality dam accidents happened in the US and other Western countries about the same time, lower fatalities owing to lower population density rather than some different level of diligence.
The solar claim is dubious (due to assuming installation of solar on residential roofs, when large scale solar would involve installation at ground level, a safer process.) But in any case, the cost of putative lives saved for the excess $ spent vs. solar would far exceed the $9 M statistical value of a human life used by the NRC to judge if regulations are cost effective. So, the argument you are making would imply a much higher value per human life, which would imply the NRC is dramatically underregulating nuclear power plants.
When managed well (i.e. Adm. Rickover's US Navy) nuclear power produces zero health-related incidents over some 14,000 reactor-years. When managed super-badly (i.e. Soviet Navy) you get multiple deadly incidents per facility per decade.