Hard to estimate long tail risks.

Chernobyl definitely didn't turn out as bad as it could have, but on the other hand hopefully we're in a bit better of a safety stance now across the industry than Chernobyl was in?

The real long tail risks from nuclear are the long term ones, which aren't factored in because we don't know what nuclear storage sites will be like in 100 or 1000 years. That's what many rational people object to about nuclear power — it involves a massive discounting of future risk, and we know people are really bad at over-discounting future risk in general.