Author here. The Mongol invasions are in the .csv in the appendix (and represented in the scatter plot), but weren't included in the table because it restricts to events that lasted less than a decade.
If you restrict to the single "worst" decade then the Mongol invasions would have been high enough to make the list, but I didn't want to start making too many manual adjustments to the data, so I left it as-is.
I think the proposal is directionally correct, but a bit too sweeping. Saying that "every paper should have a societal impacts section" is naively prescriptive, as can be seen by comparing to statements such as "every paper should have an experiments section" (false for theory papers), or "every paper should have a future work section" (in some papers, future work is folded into the intro).
On the other hand, I am in favor of the idea of holding claims about societal impact to a higher standard of rigor (although worry that the effect might simply be for people to stop making such claims altogether). I also think that encouraging (but not mandating) greater discussion of societal impact would be very positive.
Here is an alternative proposal, motivated by the often space-limited conference format used in computer science: allow for an optional extra page specifically dedicated to addressing societal impacts of the work. Reviewers are free to read or ignore this page as they choose, but crucially, this page does not count towards the page limit of the conference.
The goal of the above proposal is to eliminate the need for authors to choose between addressing societal impact and intellectual impact in a paper (even if I care a lot about the former, I know that most reviewers [correctly, in my opinion] care more about the latter, so there is a strong disincentive to spend space on the former given a strict page limit).
If you restrict to the single "worst" decade then the Mongol invasions would have been high enough to make the list, but I didn't want to start making too many manual adjustments to the data, so I left it as-is.