A scenario where the fire continued for 4 weeks instead of 10 days takes very little imagination and would be much much worse than what actually happened.
>Fukushima is a proof that a better design (albeit far from stellar) will result in orders of magnitude less damage
Fukushima would not have had a graphite fire, but we were lucky to not get a fuel rod fire which would have been very bad as well, although likely more limited geographically than Chernobyl.
As it is, Fukushima will perhaps "only" cost an estimated 500-600 billion dollars for the Japanese taxpayers over the next 100 or so years. Unless another earthquake hits and breaks the ice wall of course, which also requires very litte imagination, it being Japan.
Seriously, why are we doing things like this when we have cheaper, faster and safer options? Why are we not instead going for the options?
"It will require so much batteries" Yes. So let's get started, we can build as many batteries as we need 10 times over. "Synth fuels are inefficient", yes, but that doesn't matter when the electricity is close to free.
What else is the problem? Why does it HAVE to be nuclear?