It's a bimodal distribution, where a group of people cares deeply about the meaning a graph presents, and another group of people cares deeply about how wowing the graph is when presented.

The intersection of those too groups is too small to make any difference, thus tools will be heavily optimized for one of those.