Any country that can make a nuclear bomb could decide to make one whether or not they chose to have a civil nuclear industry (Israel being the prime example).

And in the current geopolitical climate, expect more countries to build a bomb.

If we're talking about war, what does more damage to the surroundings - dropping a nuke on a solar plant, or dropping a nuke on a nuclear plant?

Dropping a nuke on a city where nuclear plants aren't … And it's not even close. That'd be exactly like the difference between the sole victim of the Fukushima nuclear accident vs the 19 000 dead from the tsunami that caused the accident.

If nukes get involved, all bets are off no matter what, millions of people would die and the consequences of a subsequent reactor meltdown would be negligible compared to the mess you've got already.

And even compared to a conventional war, nuclear accidents are benign next to armed conflicts. (Not only during the war, but also decades after: most people are familiar with the Chernobyl red zone, but there's red zone in France due to the eternal pollution caused by WWI ammunitions).

> the sole victim of the Fukushima nuclear accident

This is a misrepresentation. There is a single person who the courts have established was (to their satisfaction) killed by nuclear exposure from Fukushima, although even that is quite debatable.

But that doesn't mean there weren't any victims, just that they could not (or could not yet) be identified. The estimated ~200 cancer deaths from Fukushima will mostly be lost in a sea of cancers from other causes. This doesn't mean they can be, or should be, ignored. Regulation is not like criminal law; one does not have to prove a technology is guilty beyond reasonable doubt to regulate it.

> The estimated ~200 cancer deaths from Fukushima will mostly be lost in a sea of cancers from other causes. This doesn't mean they can be, or should be, ignored

In comparison to the 19000 persons who died directly from the Tsunami? Yes it can be neglected. That's two orders of magnitude smaller!

> Regulation is not like criminal law; one does not have to prove a technology is guilty beyond reasonable doubt to regulate it.

No industry on earth is even remotely as regulated as nuclear industry. Over the span of the period your “200 excess death” have been calculated, more people in that particular region of Japan will have died from industrial causes, from any other industry (you should check how many people die each year from professional deceases in places as mundane as hairdressing saloons … Should we ban hair coloring?)

What nonsense. Of course we cannot ignore the 200 estimated deaths from radiation, just because people die from other reasons. You might make a cogent case that the value of 200 lives isn't all that great compared to the benefits of nuclear, but whether 19,000 people died in a tsunami is irrelevant to that argument.

Of course it is relevant: the “nuclear accident” was caused by the tsunami in the first place!

It has never been a nuclear accident to begin with, it was just a negligible (<1% in the pessimistic estimates) aggravation of the consequences of natural disaster.

Also nobody died from radiations. The additional cancer is caused by contamination, which is an entirely different health hazard for all intent and purpose.