Stephen Few critiques circular data visualizations in his article “Our Irresistible Fascination with All Things Circular“(pdf), killing among others David Mccandless’ “Colours In Culture” (interactive version here). He concludes with an example of a circular visualization that is useful (and interactive 😎 )
One of the few things that circular data visualizations can display well are the relationships between a single list of items. The following example, titled “Naming Names”, appeared in the New York Times on December 15, 2007. The circular arrangement of a single list of comparable items—presidential candidates in this case—lends itself nicely to the display of relationships between them, represented by lines that connect them.
I am not aware of any proof that McCandless work is any worse than a more ‘effective’ representation, in terms of its real impact on a lay audience. Effectiveness, here, is only argued from the viewpoint of human cognitive and perceptual science. What is assumed is that users approach these infographics with high expectations, data-driven analytical questions, or highly detailed tasks in mind (e.g. “find the color for ‘mourning’ in Native Americans”), which must be solved as quickly as possible, and as accurately as possible. They don’t. And I bet a huge majority of people still prefer the circular diagram over the matrix depiction, regardless of not being able to figure out the color for mourning.
What we should do is learn why people actually prefer the less ‘effective’ infographic, and apply this knowledge to further the field (e.g. people prefer circular graphs). What we should not do is critiquing a talented designer because a panel seemingly frames his work incorrectly as having complex analytical value.