If I may offer a suggestion, perhaps mention that pseudonyms generally tend to be long-held identities - a lot of people think that pseudonyms are just throw-away identities used to gossip about your employer safely, instead of identities that persist for months, years or even decades. (A lot of people would recognize the handle CmdrTaco, but how many know his real name?)

(p.s. your backyard is awesome)

"pseudonyms generally tend to be long-held identities - a lot of people think that pseudonyms are just throw-away identities used to gossip about your employer safely, instead of identities that persist for months, years or even decades."

A fact too often forgotten. This is the same ignorance that keeps me from using Google+ while I'm prolific on Twitter.

Honestly I don't see how you could possibly have a reasonable methodology for the claims you make in the infographic.

How do you avoid selection bias, particularly in regards to frequency of comments? For example, a source of technical bias could be that it is a simpler process to log into a Disqus pseudonym than to authenticate with Facebook when posting. Usability is incredibly important in attracting repeat users, so if it is easier to post under a disqus username then it is not surprising that pseudonymous users post more frequently. Another source of obvious selection bias is the disqus branding: I expect people who have Disqus accounts are both more likely to frequent blogs, and are more likely to recognize that they can easily comment on a blog with a Disqus comment section. By way of contrast, Facebook users may not even recognize that they have the credentials to easily comment, and may be reluctant to share their credentials with your third party service except in cases where they have a strong opinion about a piece of content.

I can see the data on comment quality being more objective and interesting, but even here there is probably selection bias: if most of the indicators of quality come from Disqus accounts' appraisals, then you may be measuring whatever bias Disqus users have as regards comment quality from Facebook commenters.

There's no selection bias here (other than the obvious, blogs using disqus), since they are using aggregate data. They make the false claim that pseudonyms contribute more though which does not follow from their data. They should show comments per user per login source data to make such a conclusion. Their current data only hint that using a disqus account is easier than facebook or anonymous.

I don't think there's such "comment racism" in disqus. I 'd argue most disqus users barely notice they are using disqus.