@schattenjaeger, definitely a discussion that needs to happen.
On the flagging front, one possibility I thought about (back when the so-called "whale no-vote experiment" was active) was turning the whole flagging/downvote issue into a two-step process. A "flag" literally becomes what it is everywhere else... an "alert" that there might be dodgy content. But maybe there's no actual "downvote" till there's some level of concensus... maybe 5 or 8 flags on a post.
The purpose being that personal vendettas and petty squabbles might be taken out of the equation, somewhat. Could they be overcome with bots? Sure... so let's say someone spams, and gets flagged. When there are six flags, the the collective weight of the flaggers then becomes a downvote by account called "system" (or something else) which also removes individual fingerpointing somewhat.
Don't know if it would work... nor if "downvoting by consensus" is a desirable approach. Just thinking out loud here.
As for self upvoting... I'll summarize my two bits (written previously, elsewhere) as this: For me there's something ethically dodgy about voting for your own content... and especially comments. That said, if I arrive "late" on a post that has a LOT of comments... and I put 20 minutes into writing a long and relevant comment? I might upvote it simply to make it more visible (as I might do with this comment... we'll see)... like people used to do when they would "bump" threads on old style message boards. I do the same thing for other people who write great comments and get overlooked, however.
That said, since I think most self-voting is done for money, I have no effective answers, there. Most people are greedy and put money far ahead of morality.