If you want to badmouth nuclear by spouting off its worst-case scenarios rather than its average or expected scenarios, I invite you to do the same for the alternatives. Coal's worst case causes the Great Smog and Beijing and New Delhi. Oil's worst case causes the Exxon Valdez and Piper Alpha and the Texas City Refinery explosion and oh yeah global climate change. Hydro's worst case is the Banqiao Dam burst, which directly killed 200,000 people and forced 11 million out of their homes. Badmouthing nuclear over fukushima and chernobyl is like badmouthing cars because a Fluorine tanker crashed in downtown Seattle around 1935.
We're essentially strangling the innovation that would fix the things we're terrified of, which are why the regulations exist in the first place!
As a result, we still have Gen II reactors (Watts Bar) coming online. In 2016.
Engineer: "The one about the primary turboencabulator?"
Bureaucracy: "Yes. Your reactor needs it to be certified to the standards laid out in Index 437."
Engineer: "There aren't any turboencabulators in the design. Our design uses retroencabulators. That's how it's so safe; the primary turboencabulator was the primary failure point. There is no primary turboencabulator to certify."
Bureaucracy: "I'm sorry, but we can't accept this design without a certification for a primary turboencabulator. It can't be safe without one. Maybe you should start over?"
If he has documented tests etc, how about going the route of getting the paperwork amended? And if not, should the bureaucrats just take his word that this time it is safe? That was what they were saying with the turboencabulator as well...
I would hope that this is less of an issue in nuclear regulation, due to them hopefully having engineering resources on staff. But I do know it's a common problem with less technically focused branches of government.
Asking for "perfectly safe" is idiotic. There's no such thing. There is always a chance that you will get wrapped up in your sheets and strangle in the middle of the night, or that you'll accidentally stab yourself in the wrist with a pencil and bleed out at your desk. Safety is relative. What we have today is better than what we had twenty years ago. What we had twenty years ago was better than what we had forty years ago.
I'd take Fukushima over coal or oil or hydro every day. It took every variation on malice and incompetence that humankind could devise to make it do anything bad, and its "anything" was honestly kind of pathetic compared to the "anythings" for anything else. I'd take a modern reactor over it, but it's take it over the alternatives. And, yes, I'd probably take whatever people have come up with in 30 years over today's technology. Science marches on.
Safety systems, backup safety systems, backup-backup safety systems can go down, and yet, the plant kills nobody. Even though the radiator had a meltdown, it was passively contained and the radiation didn't leak out in dangerous quantities.
You mention that we don't know how likely a bigger containment was to occur, but I think there's barely a nugget of truth in that. The way the radioactive material was contained is very well understood. The safety systems worked, despite anything TEPCO did.
This is the problem with the pro-nuclear approach to nuclear safety. Fukoshima was a massive safety failure. Something that was supposed to never happen did, the cooling failed, the reactors melted down, the containment breached and radioactive isotopes leaked and kept on leaking. At that point, it's not engineering that's protecting people, it's sheer luck and we don't even know the odds of the bet or the stake. We don't know how likely it was for an even bigger containment breach to occur, for things to have gone far worse, because the whole event was outside of the plan. Not only that, it was due to an issue that had already been pointed out by TEPCO's internal planning documents and ignored. (Then there's the huge economic costs of the radiation leaks.) Just ignoring this as a minor incident is inviting further failures which may not be so kind.