I think connecting the rank plots as they move through generations to time points along the time series would be a useful visualization technique so users can see what the distribution looks like as the summary stat (like a Hill #) changes. We'd use the crosstalk package for this.