I appreciate your substantive and informed comment, even though I strongly disagree.
"Voters will most likely just upvote themselves all the time."
It's been tried, and through socks it happens alla time. Remember @mindhunter though? He's just gone after his ploy was revealed. The community disapproves of self voting, and for good reason. Why have rewards at all? To create incentive to produce quality content that markets Steem and potentiates capital gains for stakeholders. While curation rewards were instituted to further create incentive to do that, they don't amount to much for insubstantial stakeholders, who in general just ignore them as a reason to upvote.
Curation rewards thus aren't any incentive for the majority of votes. For those with substantial stake, curation rewards are the end - not curation. Votes cast to attain curation rewards by those folks are cast without regard to content quality at all - contrary to the purpose of curation rewards - but solely and strategically to gain rewards.
Until vote bots arose self votes and similar mechanisms were roundly criticized as simple profiteering (and that censure has now included votebots), that at best did not create incentive to come here and buy some Steem and drive the price up. That's a good reason. So, let's admit that curation rewards don't work for the purpose they were instituted to achieve, and go on. @edicted proposed giving authors a slider, so they could set curation rewards to whatever they wanted, and despite my initial advocacy for eliminating curation rewards completely (as simply another vector for extracting rent with stake) I agree this is a better solution, as it becomes a means of attracting votes by letting voters attain to the rewards themselves - eliminating bidbots and enabling authors to promote their posts by eschewing rewards to the degree they are willing to.
As my goal for Steem is nothing short of world dominance, I want to see investors acting to create capital gains, and float every boat. Profiteering is contrary to that, despite lolbertarian freedom to do what one wants with one's stake. It isn't wise to drain the blood from a barnyard animal you'd like to sell at market for a good price, and that is analogous to profiteering and extracting rewards from the platform when you'd like to see Steem generate capital gains. Plus, it's gross, and the community does seem to dislike it.
Instead let's consider mechanisms that allow substantial stakeholders to get dividends by delegating to development projects that potentiate capital gains. Let's also limit potential rewards on posts, to eliminate that vector as suitable for profiteering. Huey Long proposed that no one should live on less than 3% of median income, nor attain to more than 300%, and I reckon that's not going to impact actual content quality negatively, as that's a significant range that allows incentive to be exerted, while also making bidbots useless for folks that abuse them just for rewards. Now, Long proposed that when he was very popular and no little ferment was ongoing in the US, and got assassinated for his trouble. I'd prefer not to get shot.
The rewards pool has been gamed to the point that Steem is suffering from poor enough optics that we underperform the market. Tweaking curation rewards isn't going to fix the problem of profiteering, as you acknowledge, and simply eliminating curation rewards won't either. Creating a vector for dividends from delegating to development projects, and limiting the extractive potential of votebots does make it possible to decrease financial incentive to abuse curation for profiteering, and enabling dividends from delegation presents a mechanism that will drive capital gains. Maximizing rewards at any cost isn't wise. You point this out:
"...not necessarily and straightforwardly all kinds of activities == incentives."
Maximizing rewards for improving development and driving capital gains is what actual investors should want. It sure is what the market wants to see, as well as most everyone on Steem today. If you got this far through my extemporaneous comment, I am grateful.
Thanks!
50/50 alone isnt that effective, from my pov the fix would need other measures as per linked in my comment above. It should be thought of as a chord to play and wouldnt sound as nice if we just play one note. If curation rewards aren’t important, then 100% author rewards would just cause more to self-vote and is the equivalent of 100% curation rewards as I’ve demonstrated above. I’ll just copy and paste trafalgar’s comment on it:-