I'm not in support or opposition to the proposal, but I have feedback...
(i) "Let's try it and see what happens" may not be the ideal methodology, and I'm not sure that it demonstrates a sufficient level of fiduciary responsibility for a blockchain with a $120 million market cap.
(ii) Mathematically, there must be a positive-value voting scheme that's functionally equivalent to one that involves voting with negative votes. Or at least one that deescalates the flag wars, rather than providing them with more fuel.
(iii) Something modeled after second price auctions might serve the dual-purposes of discouraging votes that overvalue a post (whether self-vote, collusive voting, or for any other reason), and also disencentivizing downvotes that are wildly out of step with the community.
(iv) If implemented, how long is everything going to be stalled waiting for the rewards pools to return to equilibrium?
(v) Is there any quantitative evidence to suggest that the proposal is better than the status quo?
RE: Downvote Pool Deep Dive