That logic is flawed. Some subset of instances of correlation reflect causation. A huge portion of scientific enquiry is based around establishing correlation because prior mechanistic, theoretic and empirical study deems a causative link plausible. Would you wish to do away with the collection of data on correlation? My comment was made in a context in which a causative link is clearly plausible (that American foreign policy makes people angry). Therefore quantitative study of American foreign policy actions and instances of angry foreigners would be a perfectly valid exercise. "Correlation is not causation" is a gentle reminder not to make facile inferences found in undergraduate textbooks -- you appear to be under the impression that it's some sort of death blow to scientific epistemology.