Since it's rank-based, it would be nice to see a comparison to Spearman's correlation instead. It's very easy to find failure cases for Pearson, so the visualization is next to meaningless, IMO.