It's important to note that "extremism" is defined in this experiment by the intensity with which one holds a view (according to their own response on a 7-point

It's important to note that "extremism" is defined in this experiment by the intensity with which one holds a view (according to their own response on a 7-point scale between "strongly against" and "strongly for"), not (for instance) by how uncommon or unpopular a view is. I would be curious what results they would get for much broader political questions with the potential for more "radical" (i.e uncommon) results.