I highly disagree with a separate pool for downvoting. You can't expect the platform to get better by cutting decentralization in half and giving whales twice as much power.
I also suggested non-linear rewards the second I touched down on this platform 10 months ago. I had no idea it had already been attempted and abandoned.
At one point I also came to the conclusion that 50% curation was a good idea. Now I realize that the curation lottery is an easily gamed system as well. Anything with set rules will easily be gamed.
The system we have now works perfectly. It is the simplest most straightforward option as the foundation of the platform. Introducing these new "solutions" to the problem will only allow the system to be gamed even more by those who understand the complexities of it. As you already know, whales are highly motivated to do just that.
The solution isn't in changing the foundation of the Steem blockchain, it's building on top of the anarchy to create order.
The single most important thing that this blockchain needs is a decentralized reputation system. We need a way to organize big groups of people that all share similar crypto ideology. We need to be able to measure who is trustworthy and who is not. This is all that matters: trust.
None of the solutions you've presented will force anyone on the blockchain to be more trustworthy. Unless we can come up with a rule that's as solid as POW or POS or DPOS it isn't really worth hard-coding into the foundation. :(
Our proposal atm is a 10% separate downvote pool, a contingent to the other two measures which is 50% curation, and n^1.3 reward curve. Gamed or not, n^1.3 forces congregation of posts with substantial payouts and makes pricing of votes less predictable than linear. It's a more manageable playing field for any regulation to take place. Nothing can be worst than what we have now as content-indifferent behavior gets rewarded the highest. In the end, it really depends if you agree on the problem statement and solution.
https://steemit.com/steem/@edicted/steem-problems-can-t-be-fixed-with-hard-forks
Now is not the time to be changing the core dynamics for how the platform operates. Complicating the rules even further will simply reward the users who are best at gaming the new rules. It most certainly will be worse. It will be even more centralized.
You base everything on the premise that self-upvoting is the most profitable action. This is false. Working together to give the platform more value is the most profitable action. These solutions will not promote synergy. It will be more of the same.
I would implore you to read that post where I directly respond to your proposal. Occam's Razor is our friend. Adding more variables is counter productive.