I also try to keep my points more limited, e.g. I cast doubt at is your claim that we know the risks well, rather than settle down for a figure or a comparison with the risks of a rival technology. *Lunatic*, n. One whose delusions are out of fashion.
That makes it very hard to quantify risks accurately.
If the Windscale fire problem hadn't been solved, it's likely that much of the Midlands would now be uninhabitable. You can more or less get a way with an accident like that in Russia. In Europe, population densities are somewhat higher and the prospect of losing large sections of a country aren't appealing.
So it's not just imputed fatalities, clearly. The overall risk has to include the extremely small - but not quite zero - prospect of a more serious accident.
As everyone else has been saying, the sanest way to assess the risks is to look at governmental competence and corporate culture.
From that point of view, nukes are politically demanding. If insurers don't want to insure them, it could be as much because they know that the political stability and oversight required to minimise risks isn't there, as because they're worried that any trial will be railroaded by hordes of angry chanting hippies wearing sunflowers.
