Exchanging ideas? No way. Interesting viewpoints, if they stand out at all, will manifest as odd data points that are indistinguishable from user error. This

Exchanging ideas? No way. Interesting viewpoints, if they stand out at all, will manifest as odd data points that are indistinguishable from user error. This is only useful for studying large groups of people in aggregate and characterizing them according to ideas already understood by the experimenters. I.e., it's just a fancy way of gathering and visualizing polling data, and it doesn't seem very original. What are the odds that political consultants haven't already applied PCA to polling data (and probably been disappointed with the results)?