I disagree with regards to flagging to reduce potential post rewards. As you mentioned, there are numerous examples of posts earning a vastly disproportionate payout, and we know what they are: spam, bot abuse, plagiarism, and the like. However, when it comes to genuine disagreement over the reward v. the content of the post, a flag in this regard negates the stake used by other users to reward the author. Functionally, I know that's not how it works, but ultimately that's what it comes down to. User A upvotes a post and brings it $12, let's say. User B comes by, takes a look, and flags it, reducing the net effect of that $12 down to $3. In effect, User B is telling User A, "no no, you made a mistake. Here, let me fix it for you."

It can be called reward pool reallocation or any other term, but that's essentially what it boils down to: "I know better than you what deserves these rewards, so let me fix your mistake."

My suspicion with regards to the whitepaper is that the intent was to guide users to downvote blatantly cheap posts to fight upvoting collusion, i.e. three-line posts upvoted by a group of people to guarantee curation rewards. I could be wrong, and @ned should clarify the intent of that particular guidance, but that's my impression after being here for more than a year.