There was a lot of antisocial and other atypical behavior going on between March and May that year.

Yes, that's exactly why I expected more influence on the activities at this time. The fact that a few more accounts were created there is the lesser evil.

unless someone wants to scan the blockchain or hivemind for author reward transactions.

I also used your burned token posts for the estimation. :-)
Too bad that we didn't have the data saved somewhere. But it's not worth running the chain because of that.
I have discovered an alternative (paid rewards). I looked at an account that posts daily (nodex).
With this query and the appropriate times, we can determine the reward shares (and thus also the print rate): https://sds0.steemworld.org/rewards_api/getRewards/author_reward/upvu.proxy/1598918400-1656633600

As you can see, there were fluctuations in the print rate at the end of 2020, but after that it was fairly stable until May 2022.

SteemOps

Have you downloaded the data? I'm a little hesitant because I don't know how large the amount of data is.