You're conflating possible with probable. Many things are possible with a probability approaching infinitely improbable. Something catastrophic and highly improbable already has happened, and three years later still approximately 0 (zero) deaths from radiation. Probability theory suggests the chances of another catastrophic event occurring at the same site are much smaller than the first set of events.

I don't want to nitpick but probability theory doesn't care that the event already happened at the site. So presumably the risk of another tsunami coming in or a major earthquake is pretty much unaffected by the fact this specific site already had seen a major disaster... I guess the site itself has changed and that means something in the terms of the probability curve, e.g. it can't get much worse or something along those lines...

Earthquakes are caused by two continental plates releasing tension as they slip past each other. Earthquakes (and therefore tsunamis) are not independent events, and strong quakes are unlikely to occur in close succession.

Probability theory defers to physics.

As someone who lives in an earthquake zone I think this theory isn't consensus by any means (that earthquakes release tension and reduce the probability of a followup earthquake). That is no one actually knows if having a strong or a weak earthquake increases or decreases the probability of having another one. Once we can dissect the earth and model accurately what's going on in there it will become physics, until then it's more like geology... Tsunamis can also be triggered by earthquakes in different locations.

I'll agree with the part that earthquakes aren't independent events...

EDIT: "The opposite occurred in Turkey along the great North Anatolian fault. A large quake in 1939 sent stresses farther down the fault, triggering a 60-year series of quakes whose latest installment was the deadly Izmit earthquake of 1999. The stresses have risen in the crust near the city of Istanbul, and a quake there is now considered more likely. " - http://geology.about.com/library/weekly/aa022303a.htm

I think you are confused. The earthquake and reactor accident happened.

That the reactors are not contained, is not a separate event. It is cleanup that has not yet occurred.

You wouldn't say, "Well, I fell off my bike at 25mph on pavement, landing precisely on both kneecaps, which was highly improbable" and then say "the probability that my knee is bleeding is therefore lessened", would you?