It's not pseudoscience if there has been an empirical study, almost by definition. You're right that the value of self-reports is pretty low, but there have been plenty of studies done similar to this, this is not a new concept.