Very nice paper and interesting result with respect to angularity. Almost paradoxical as round is more prototypical on average, and prototypicality correlates with beauty, but round is lower in rated beauty than angular. The speculation that this has to do with rating vs forced choice is problematic because round objects are judged more beautiful in studies that used ratings and abstract shapes not dissimilar to flowers (Bertamini at al 2015, Palumbo Bertamini, 2016). Indeed the preference for smooth contours is quite robust. Maybe one explanation could be that all angular specimens had radial symmetry (from what I could see in the dataset), and that bilateral symmetry was judged less beautiful than radial? By the way this is consistent with the increase in beauty with number of reflexion axes, although this increase is not monotonic.