Interesting analysis @crokkon, although the results are not entirely surprising. I'm glad you mention @dustsweeper, as it is one of the very few "countermeasures" those of us who are active posters and commenters have.
This does make me wonder about the efficacy of some curation trails; one can but hope that the many people who are getting a 6% upvote from a series of accounts with 87SP (or something like that) aren't just getting "a nice warm feeling" but getting NO actual rewards.
"Margin of error" matters too, here. Meaning that having a 0.020 value right off doesn't mean it will still be 0.020 seven days later... which is also where the DustSweeper "overvote" comes from
Thanks for "spelling this out," however!
=^..^=
.