I've thought about improving curation reward but it's very difficult to incentivize people to vote for what they like instead of what the post will earn.
A few months back I suggested considering "time account spends on X post" as part of the voting calculation.
So the voting calculation could be something like:
(SP of account, factoring in current voting power) * (time spent on X post) * (1/0: upvote or not?) * (voting slider value: this is where we would state whether the read was worth the time spent)
...and we can give every X time interval spent on a post some incremental value of 1, until the time limit is reached (we could settle on 5 minutes, or 300 seconds, as a 100% vote by time and split up the measurement in 5 second intervals, such that spending only 5 seconds viewing an article only casts a 5/300 {1/60} time weighted vote in the above calculation)
This way people are somewhat punished for voting on content that they don't like -- they would have to spend at least 5 minutes with their browser stuck on a post that they don't like (or 10 minutes, or whatever value we determine to be the time limit) in order to cast a 100% vote on it.
If we could somehow code accurate account viewing time per post where the first post opened is the ONLY post on which the time measurement happens (UNTIL that post is voted on or the post is closed out of), then I think we would make gaming curation through bot-voting much more difficult to achieve AND better incentivize voting on things that we like vs. on things that we believe will be popular with others.
So this would reward people based on the time they spent reading it? And you would calculate the average of all people who viewed the post to determine its popularity? Is this correct?
How does this work when a photographer or an artist shares a few images? A comic doesn't take long to read.