Update 2 (July 15, 2014): New, more accurate data is here.
Update (Jul 1, 2014): This post doesn’t measure releases in 2-week totals, which turns out to be a huge deal in many, many cases. I’m currently working on an updated version of both this and the 2011 data. Just be aware of that before citing the data from here regarding any one show.
By all rights, a 30-series sample like the one I had for 2011 was enough to get most of the relevant information regarding how anime boosted manga sales. However, during that analysis, I bumped into an incidental correlation, myanimelist ranking versus gain in manga sales, that was far too juicy to ignore. If that correlation is real, it points to a very tangible link between the Japanese mainstream community (who have enough disposable income for manga but not for anime) and the English-speaking online community (who generally pay a comparable pittance, if anything, for the anime they watch). But I couldn’t be sure from just the 2011 data, since that was the sample that gave rise to the theory. So I did what any good researcher would do, and pulled another year worth of data to see how things would match up. The results can be found on this spreadsheet, and are sorted in order of descending myanimelist rank below.