I was underwhelmed at the end of the article when it becomes clear that another interpretation of the results is "People presented with shocking results tend to scrutinize them."<p>A lot of Kahan's research can be read as "polarization dooms us all, careful discussion and thought are useless." I think you could find the opposite conclusion here, that communities with strong adherents of competing views are more likely to have some members discover errors in hard problems that might otherwise go unnoticed.
Interestingly, communities like HN and LessWrong have banned political discussion, or tried to, for its particularly deleterious effect on conversation.<p><a href="http://lesswrong.com/lw/gw/politics_is_the_mindkiller/" rel="nofollow">http://lesswrong.com/lw/gw/politics_is_the_mindkiller/</a>