> What do you think happens if you send a missile to the "wet surface impoundment" that releases the contents of the lake into the town or the groundwater?

I can't speak to that. But nearly 50 years ago I did a deep dive into what would likely happen in the event of a nuclear "accident" (a term of art) in a Navy ship's reactor in port. This was when I was doing the Navy's prep course for the [chief] engineer exam after two years of sea duty running aircraft-carrier reactors. Current-design civilian reactors are much larger, so the effects of a missile-strike meltdown would be correspondingly worse. If I had to guess, it'd be far worse than even the missile strike you postulate.

Footnote: AFAIK there has never been a nuclear accident aboard a Navy ship, submarine or otherwise. That's something in which nukes take immense pride. It's largely because of the zero-defect, second-checking culture ferociously instilled by Admiral Hyman Rickover during his decades in charge of "The Program."

Back to non-missile dangers: Human error is what I've always worried about for nuclear power plants. From what's been made public, both Three Mile Island and Chernobyl could easily have been averted — had it not been for cascades of operator errors. Can we confidently say that such errors are less likely today? To be sure, many civilian nuclear plants in the U.S. are run by Navy veterans. But my guess is that working in such plants doesn't provide the same motivations and incentives as "the Fleet." (And a flock of suicide drones won't care either way.)

Relatedly: I was just reading an account of Air France Flight 447, which flew itself into the middle of the South Atlantic — killing all aboard — because of cascades of egregious pilot error that defeated all the autopilot systems.