In complex systems accidents do not tell the whole story, one has to also look at incidents and the "safety culture" of the organization (i.e. ~ how they avoid failure and how they react to failure).

Aviation for example has The Aviation Herald, where one can see that there's in fact quite many incidents that almost never result in accidents. There's lots of literature on safety engineering in aviation.

What's the equivalent for the nuclear industry? I was admittedly not very interested in this topic, so I'm only aware of the descriptions from Perrow's "Normal Accidents" book where he reported on several incidents which were actually published by the involved organizations. They didn't inspire confidence at all, on the contrary they were downright scary (and much scarier than one can find on the Av Herald).

Maybe the nuclear industry grew up, but it's absolutely necessary to have transparency here. Looking at global accidents is like trying to map a castle through a keyhole.