Not only are counterterrorism operations shrouded in secrecy, but it's impossible to prove or disprove claims that enhanced interrogation works better than other methods when prisoners are intent on saying nothing.
Scientific study of this question would require random sorting of suspects into groups that receive either torture lite or conventional forms of interrogation.
To frame this inquiry is to show why it can't be carried out: It would violate international law and research ethics. The CIA, Hubbard told me, conducted no such study for this reason
So, torturing people might be legal and ethical (according to the CIA and their pet psychologists, anyway). But randomly selecting some subset of prisoners not to torture would be clearly illegal and unethical? Is it so hard to get informed consent to be not-tortured? I mean, clearly longstanding international agreements don't mean anything these days, but violating ethical constraints on human research--- that could get your funding pulled.