Interesting idea! Would be nice to see:

* How the colors were picked and assigned to each category and (e.g. at what point is red pink and no longer red)

* An indication of distribution in charts, they have different scales on the y-axis.

* The author likely sampled posters with mostly the same color above a given threshold for each category, would that (together with the lack of methodology and error bars) heavily skews the reader's presentation of the data analysis?