Surely the most logical way to test this hypothesis would be to write a normal comment and write a "provocative" comment and see if there is any difference between the two comments made at the same time on the same account. Had the author done this they would have found out there was no difference and this is simply down to the public parts being heavily cached.