These aren't really worth cluttering the board on their own but are probably worth doing at some point:
[x] Displaying points when the range of sample sizes is large helps avoid comparing incomparable things (since small n -> more variability). Some alternatives we'd talked about were: including the sample size in the hover text or as a label, or including sample size in a table off to the side.
[ ] to make the exclusions easier to use, we discussed only applying an exclusions list when a "submit" button was pushed. This avoids re-rendering plots for every single change in the exclusions list.
These aren't really worth cluttering the board on their own but are probably worth doing at some point: