Minor update to the SNM paper: we discovered and fixed a small rounding error in the enrollment-balancing metric code, which had a small quantitative effect (and even smaller qualitative effect) on the section-size distributions for the 2008-15 cumulative results. So, we've released v3.1 on the SNM website, updating one figure and two tables. (And, the '-current' PDF now points to v3.1.)