A linear curation reward curve would remove all incentive to vote on good content (for the purpose of curation rewards). If your vote is linear then it will increase the payout of a post by the same amount no matter how many or few votes the post has.Here I don't agree. Actually I think the opposite is the case!
Now many people just vote on stuff because they expect many others to vote on it as well, for example because an article is written by a "famous" author who always receives many votes. That means the voting process often is more about speculating what others may vote for, instead just to vote because one really likes an article.
If one would receive the same reward for every of ones votes, then only people would start to vote for stuff they really like.
I intentionally try to vote for articles with only a few votes, if I feel they are good, even if I know my reward will be smaller then.
No, a linear curation reward curve would kill one of the main ideas that I have about Steemit. Which is to surface undervalued content and getting rewarded for this effort. If the curation reward is just a static increment, then there is absolutely no incentive to search for anything yet undiscovered.
I see that differently (probably also because I earn more by writing articles instead of rewarding them). I believe you that sometimes you really get rewarded by discovering undiscovered stuff (I know that you are a very good curator), but I think often curating is more about to reward stuff of famous authors at the right time.
Of course there are curators like you who seek "hidden" stuff, but I have to disagree that articles are "good" only because they have many votes as @vandeberg wrote. I can conclude that also from my own articles where I often have more than hundred votes but only lets say 15 views. How can the 90 % of voters, who didn't even read my article, decide if it is good or bad?
Another example: let two persons, one of them being a "famous old" member, and the other one an unknown "minnow", write exactly the same article: you will see that the well known user has many votes within a very short time, whereas the unknown user has only a few votes within a long timespan. I will bet on that.
This observation is correct and it´s exactly the reason, why reward-oriented curation tries to spot content from unknown users. Trust me, try to maximize your return on curation for a while (if you can afford spending the time that this requires…) and one of the first things you will stop doing is voting on high-rep authors. As you say, the usual suspects of the trending page collect rewards so early after posting that there is simply no window of opportunity for performance oriented curators. Therefore I say, incentivizing content discovery actually counteracts on this kind of inequity that you find faulty here.
Thanks for your interesting thoughts!