On a slightly different note, I once read this research paper talking about analysing “collective cohort difficulty” using such SRS systems. That sounds quite intersting. If anyone ever comes up with such an add-on for teachers, that can be a good source of teaching decisions.