The table seems to be sorted by user count which is probably a more accurate measure of popularity at the current point in time (huge changes in the last month, so anything measured “per month” probably suffers from some anomalous values depending on how you extrapolate the monthly figure from the time significantly under a month that has passed since the Reddit migration).