I'm quite sure, but not 100% that the above number does *not* include research reactors. If it makes sense to include those into the statistics is less clear.
Fukushima had three meltdowns (reactors 1-3), and a hydrogen gas explosion that damaged the spent fuel pool of reactor 4. Together with Three Mile Island and Chernobyl, this corresponds to roughly 1% of all commissioned reactors suffering catastrophic failure with significant radiation release.
You both are making statistical errors. You've aggregated the data improperly. As I mentioned in the original rant, this calculation throws out the temporal component and thus 80 years of development and research. Safety has greatly increased over the last 20 years, let alone 80.
I'm quite sure, but not 100% that the above number does *not* include research reactors. If it makes sense to include those into the statistics is less clear.