Human error makes sense and I guess that falls into a kind of third area of risks really. Avoidable and prepared for but still happens a lot because of people trying to take short cuts or not following procedures.
The issue with the nuclear plant and the deterioration of the core doesn't fall into that category. It falls into a category of monitorable plant deterioration, can and is modelled using measurement and inspection.
The other category is unexpected hazards like flooding, tidal waves, earthquakes, terrorists. Can also be planned for and try to prevent but there is always something new to come along and bite you I think in this category.
What I mean by my initial statement is this... where is the counter challenge? So for example an experienced nuclear engineer comes along and identifies cracking in the core. He (his team) measures, models and evaluates the best they can and come up with a recommendation. Their recommendation will always be on the cautious side, perhaps rightly so. But where then is the challenge saying what if...what if we could extend. What are the options, mitigations, extra inspection and measurement regimes we could implement and how much does that change the risk? Maybe they have done this, but its not transparent that they have in this example.
Do you not think risk assessors are conservative? Isn't the main idea to prevent first, then mitigate if prevent isn't an option?