I don't believe changes such as these will alter the behavior of large stake bad actors in any way. The main problem is that these accounts have large votes that single handedly propel the rshares of the content they vote on to a very high point in the curve which matches the current economics, thus no change occurs for them. There also will be no change to self voting patterns because this remains the best ROI for them. Even with a 50/50 split enacted, large accounts would have to have votes nearly equivalent to their own voting strength come on to the post subsequent to their upvote in order to begin to match the gains they'll get from self voting and scooping both the author and curator payout on their own content. This doesn't happen because of their sheer size, as it takes a huge number of small users to match them. The only viable alternative for large accounts is to split their vote into many smaller votes, which I'm sure is the preferred outcome, but is entirely unreasonable to expect we'll see better quality curation out of this. Telling an investor that to see good returns they'll merely need to read, evaluate, and vote on a couple thousand articles a day using 1% of their upvote strength is a non starter. It'll still be sold, outsourced, etc resulting in no change to curation. The only thing that will begin to make curation more viable is a much larger userbase and a broadening middle class. These proposals actually stunt the growth of both with lessened author rewards and curtailed voting power at the low end. Curation should be organic. Vote on what you like. All of these changes seem to aim at making users guess what will be the most rewarded, which isn't organic discovery.
We have yet to see how the current system will function at full scale. I firmly believe that when there are tens of millions of active users and voters, we'll see genuine opportunities for profitable curation under the current ruleset, and self voting insular whales will merely handicap themselves by not growing a fanbase and interacting positively with the broader community.
The only thing I would support is some form of continuing to normalize and encourage downvotes. Much of this could be simply educational and psychological, not even on the blockchain level. Badges and avatars, a quick hover over a user that'll show stats on how broadly they curate, how much they self upvote, etc. A complete overhaul of the reputation system that embraces this. (self upvotes decrease visible rep score?)
All in all, far too much time is being spent on how to influence a few dozen people to not be short sighted... and far too little is being spent on how to prepare Steem for millions of users. We should be asking ourselves how proposals like this impact users holding 1-5 Steem power. If we get to 100 million accounts... that's how much 99% of those accounts will have. If we get to even 10 million active users in the next couple years, given Steem inflation we're looking at each account earning a median of approximately 3 Steem per year. And that's the median... since early adopters will vastly outperform that, virtually all average users following any kind of widespread adoption will see less than 1 Steem per year in earnings. Period.
Forget trying to figure out if 99.9% of the future users will get .8 Steem or .6 Steem a year. Leave improved curation and community specific rewards up to SMTs, Oracles, and 2nd layer systems as has been the seeming plan all along.
Create the much needed Resource Credit pools for dApps, communities and large scale adoption. If you want to change the current dynamic somewhat, do not allow RCs and SP to be separated during delegation. Serious businesses that have need of both the RCs and the SP for responsibly building and rewarding their communities will inherently pay more for these combined delegations than a bid bot project that has no use for the RC component and we'll see more SP shift to sustainable and responsible business applications and allow a better avenue for passive investors to pursue ROI. And finish polishing up that Promoted Post system to wrestle back some of the "exposure" spending from bid bots and create a counter to Steem inflation!