Note that they don't use 100 as the Chernobyl deaths estimate, they use 433. While I think that is very optimistic, there are very valid concerns with estimates made using the linear no-threshold model (which they discuss). And they used 2314 for Fukushima, which Wikipedia [1] agrees with and ascribes largely to the evacuation efforts.
I'm not claiming it as a 100% accurate source, as it was just what I found with a quick google - I agree using a range of data points for each energy source would be more rigorous. Do you have a better source that uses ranges for the different energy options?
[1] https://en.wikipedia.org/wiki/Fukushima_Daiichi_nuclear_disa...