I've seen some discussion on "working memory" and "spaced repetition"
I just read this pop-science article in which a new hypothesis is presented that seems to provide better predictions and test conditions for measuring working memory. Maybe this can also be used for the SRS contest.