I will further look into @sneakyninja.
Basically, nobody, including the contributers of upvote proposals, can perform an above threshold search. Each one samples more or less randomly for ten or twenty minutes and what they get is as good as anything they find in an hour or two or three. They start looking in different places.
The Daily Sneak seems like a good project.
By having many individuals sample the stream of content and submit individual upvote proposals, the variation in which content is seen and evaluated is much larger. In this case, it's that of the sum of as many random variables as curators, not just of one. And it's multiplied by a constant, which is larger as the diversity of categories and accounts where different contributers start looking is greater. How much that translates to a higher probability of finding quality content, I would have to think about some more. It depends on how quality content is distributed, clumped on Steemit. Which is an unknown factor. The less clumped, the more probable it is for sneakyninja to upvote quality content missed by most users. If he's having some success, it's probably not too clumped at the moment. The fact that voting is then done by a larger account, an all at once bump, is good.
Maybe a hundred Daily Sneaks, operating uncorrelated, then comparing approved proposals, before voting around the same time, is going to be good enough for higher quality to appear to regular users. That will bring more people in, and with more people seeing quality content, more expensive to spam unlimited.