Repectfully, you are completely mistaken about downvoting. I understand this is not necessarily a popular point of view because downvoting can cause a lot of problems. However, ultimately downvoting is the only way to differentiate between 'good' content and 'bad' content or between value creation and value extraction, or any other distinction you want to make (and we should be well aware by now of the critical importance of making such distinctions).

Curation rewards (and increasing them) are okay, but ultimately they too can be captured, sold, and redistributed. Current bid bots just keep their curation rewards, but make curation rewards bigger and they will start sharing those back with delegators and/or customers in a new form of mediated self-voting.

Non-linear rewards similarly incentivize concentration of voting (which has at times been also identified as a villain to a healthy Steem economy), but they still can't differentiate between good concentration and bad concentration. Again, this can also be captured and redistributed. For example, most current bid bots divide their vote among all bidders. But bid bots could do something different, like only vote on the top 5 bids, ensuring higher concentration and better returns. Or impose other rules enforcing sufficient concentration. It also rather directly encourages vote trading (of the form: "We both vote for all of my posts, and in exchange we will both vote for all of your posts.") This in no way improves the content or the economy. It has some modest benefits in increasing the visibility of rewards being paid out and discouraging spam, but beyond that, not much help to the quality of content or value-creating activities.

Your comments about economic incentives and getting a better equilibrium are spot on, but the way that such an equilibrium is defined to favor a particular kind of activity is almost entirely via downvoting. Our current problems largely derive from insufficient downvoting: As you said, the dominant part of the current equilibrium is content-neutral, but again, how do you encourage an economic equilibrium that is not neutral. The only answer I see that actually does that is downvoting.

That doesn't mean there will actually be a lot of downvoting going forward (particularly after a transition period, as people figure out the 'new rules'). Downvoting is costly to at least one and arguably both sides of the deal (upvoters and downvoters) and naturally an equilibrium will tend to discourage it (just hopefully less so than now), but the resulting equilibrium will have voters who are more careful about what they vote on (doesn't that sound nicely non-neutral to you?!) because they face a real risk of downvoting if they engage in bad behavior (defined above, and by you and others very well in the past).

I know downvoting is controversial and in many ways unpopular, and that it has real costs in terms of potential for conflict, discouragement, etc. But if we want a better economy and better functioning voting system we are going to have to swallow the bitter medicine, and not by watering it down and describing what is needed as being only 'moderate' (as you unfortunately did above).

FWIW, I don't believe much non-linearity is needed, in part because downvoting already introduces non-linearity (your return is no longer linear in your stake/vote power if you face a real risk of downvoting). However, I would support some modest non-linearity at the low end as an extra step in discouraging spam.

Also, one last comment. @lukestokes is absolutely right that step one in making changes like this consists of one of two things:

1. Convincing Steemit Inc. to change its development priorities to addressing it.
2. Undertaking a more comprehensive effort (as @lukestokes has described in some other comments) to de-centralize the development process and control over the code base. As things stand now, witnesses only have the ability to veto hard forks that are proposed by Steemit. Witnesses can not enact any changes that are not implemented and released by Steemit, absent a more comprehensive effort to rework the development process and control over the 'official' github repo.

I fully appreciate that your post may be, in part, directed at #1, and that's perfectly fine (even something I wholeheartedly support). However, many of your comments focus on the witnesses and unless you have #2 in mind, witnesses have a very limited role to play on this.