It's an interesting idea but the real problem is both the option we talked about earlier and the one presented here require extensive development to just test out as a concept.
If made, this wouldn't be so much used by our quality content creators or standard bad actors but by the types who collude with/manipulate others for their own ends. Honest people who received a downvote they're confused by wouldn't use it. I've never once seen a honest content creator actually ask for their downvote to be countered; they only want to know why. It's not about the monetary value for them but about the fact that someone downvoted their work.
Developing a quality of post algorithm would be a more worthwhile investment of time and money to support the PoB concept.
There have been times when a large stakeholder would, in essence, downvote every single thing to the ground; I know of people who had even their comments hit every time. As we move into the future, this becomes less of an issue, but since PoB still relies heavily on HIVE atm, it still can cause a stink. It's hasn't really happened in a while on that scale. But since the talk of PoB has come up recently, I wanted to take a swing at putting some ideas out there. @smooth pointed out something I missed somehow, the fact a large stakeholder could simply split the stake then use the FD on the other account.