The upside about the curation pool is that you can cross check between the other curation accounts to see if someone is trying to abuse the pools. The way the initial upvotes are being given can always be altered if deemed so (POB). That being said the end goal is of course a healthy economy where content needs to be consumed before it can be rewarded.

A smooth transition to the 500K in combination with letting users determine quality content might be a step forward. topic could be determined on their interaction (upvotes, comments, posts, etc) about that topic (tag) which the curation account could take in consideration (I know a lot about video games, but don't ask me anything about gardening, I don't feel my weight in the latter should be as big as in the former.) Some sort of reputation per tag...

A bit off topic in regard to POB

I've always wondered if there is any added value in adding extra parameters in how your vote value is being determined.

Something in the line of a penalty to your vote weight depending on how often you've rewarded that author in the past

something in the line of :

penalty is 1
Treshold = 0.05 (5%)

count past votes: a correlation between vote power and vote weight given to that author as well as how long ago they where given

(you can reduce the weight of a vote depending on how long ago it was)

if" count past votes"/"total past votes" exceeds the treshold add an amount to the penalty in correlation to the amount it exceeded

vote reward = vote power * vote weight * vote value (as we know it now, with vote power and vote weight both being a double between 0 and 1) / penalty

I've given a treshold of 5% here as an example. In practice this should probably be a variable number taking certain parameters in account, as the amount of active publishers in the past 7 days.

This way you reward the people that go out there to find new content and you prevent self upvoting in big quantities. By reducing the quantity of self upvotes/circle upvotes you'll also reduce the articles published and as the quantity reduces the overall quality of the articles will improve as people have the time to put more effort in their work rather than trying to write and publish multiple articles a day. Further, the people that do go out to find new content can claim a bigger part of the reward pool.

This is purely theoretical and the tricky part is finding the correct numbers. I've not given it as much thought as you have of course, nor do I know all of the current parameters. but I thought I'd share it nonetheless.

At the end of the day the closest you can come to BOP is making "cheating" so unrewarding that people won't bother. And if that isn't enough of a challenge it should feel like "non cheaters" are "rewarded", rather than "cheaters" are "punished". Many challenges ahead but topics like these are always a good way to theorize on the matter