There is superlinearity in curation rewards now, it's just not based on stake.  Essentially, the superlinearity is based off being the first to spot good content before others find it and vote on it. The early voters get more than the later voters. The idea is that this will encourage a curator to find the content that other people will vote on. 

But this is often misunderstood right now: many voters think they get more for voting for a post that already has a lot of votes on it, when such votes actually get much less of the overall curation reward. 

Most importantly, under 25% rewards split among all curators versus 75% going to the one author, this superlinearity is pretty much moot because the rewards are just so small relative to author rewards that the economics are totally skewed to self-voting. 

The most important thing to do now is to change this percentage to allow for meaningful curation rewards. Then we can consider tinkering without how those curation rewards are apportioned after they become numbers that matter.