The title to Erin Griffith's Pandodaily article on curation speaks more largely than it realizes. Why? Because the scale problem in curation is much the same as the scaling problem in MOOCs. How do you assess what folks know when there are 100,000 of them? In MOOC's that has been answered two ways: roboassessment and peer assessment.
Nothing else scales. But it appears that curators are finding a way with with what amounts to curatorial boards, groups of experts who work their way in, around, and among the participants like Wikipedia angels amongst the trolls. I think we would do well to see how curators thrash through this problem. Maybe we can get some ideas from them.