What we want to do is to attract the discovery of undervalued users and increase their popularity and because of this, a variable split that is performed automatically based on prior rewards might make more sense. In either case, we want the percentage of curation rewards to be lower for popular content and higher for unpopular content in order to encourage the discovery of meaningful content. Note that these ideas do not encourage upvoting bad content, but rather discourage the centralization of rewards among the same group of popular users and encourages users to find content meaningful to the community by undiscovered users.
What about the idea of not making the author/curator split the variable, but somewhat de-linking the author and curator rewards for a post and looking across posts to determine curator rewards. Posts from "already popular" creators would get the same type of per-post author reward, but the amount of reward allocated for curators on those posts would be low, and the amounts of curation reward for posts by "previously undiscovered" people would be higher. It seems to me that you wouldn't want to "punish" unpopular creators by lowering the percentage of rewards they get for their posts since the prospect of getting rewards is an incentive to create stuff in the first place and getting a smaller slice of an already small pie could work against that. But it makes sense to me that there should be a greater reward for identifying and popularizing under-seen content and maybe less reward for saying "yeah, that person who everybody knows posts popular stuff posted some more stuff that will be popular".
Yeah, I get where you are coming from. The above proposal may make it appear that newer users are being punished for not being popular. The issue with looking across posts would be an implementation one as different posts have their rewards due at different times, but it is an interesting idea.
We could also leave the split 75/25 and simply burn some portion of the curation rewards from popular users as they get more popular. This achieves the same effect as decreasing the percentage of rewards without affecting the author's payout.