Regarding the question of coin-weighted ballots are supposed to be tallied/interpreted, there’s a fundamental question here, that affects future protocol design.
Ideally, should the voting mechanism expose only the final tally, or also the individual ballots?
I was under the impression that we’d only look at the final tally (as ECC did), and therefore ideally should expose no more: both to maximize privacy, and also because the finer-grained information can be faked anyway. And we can cryptographically achieve this.
But this would eliminate the visibility of the small coin holders. Which @zooko appears to believe is achieved by the present mechanism (despite the lack of clarity on how to interpret such voices, and why to believe their authenticity when Sybil attacks are trivial).
So what are we even trying to achieve here? @amiller, @zooko, it would really help move this discussion constructively if you’d stick your neck out by concretely defining your desired voting/polling/petition/signalling functionality, including what information should revealed and how it should be interpreted. (You can assume away all implementation issues, but not inherent ones like Sybil attacks.)