I haven't read your suggestion, I'd like to check it out.
I'm not a big fan of what I call "bribe-a-bots". They artificially curate posts based on "donation", rather than quality content ultimately only serving the bot owners and delegator. Big upvotes go to those with the biggest budget, rather than to those who actually put forth good content.
My idea for the "bribe-a-bots" is for posts to pass a HUMAN curation audit first (according to predefined parameters) before having the bid be included. This will allow for bigger payouts for good quality posts and refunds for the garbage posters. The criteria doesn't even need to be extremely high, but divide between authors that put forth genuine effort vs lazy and plagiarized content.
Bot owners would loose business, but they would also LEGITIMIZE their business for the long term. The steemit community and development team will correct for this at some point and the paid-for bot upvotes on garbage posts PROBLEM will be solved. @ironshield
This is of course a good idea, and at least qurator vets users first.
I agree with what you are saying, but it's kind of the fundamental problem steem has already: larger stake = larger rewards. Maybe the problem is that the influence is linear in the amount of stake you have... Maybe it should follow a sublinear model...
As long as the market makes this profitable the problem remains. I'm taking the stance that I can use this to push rewards around differently. Compared to whales and abusers I doubt it's much of an impact though.