You are making statements you cannot support with facts. Test procedure does not tell us the things we need to know / have in order to know if testing is in fact accurate. Where and when is the air temperature being taken? If it's to the air temp into cooler/radiator and taken at same time as CPU temp the delta is likely to be way off. OC3D data does not tell use when, where and at what time it was taken relative to CPU temp reading. So what is his delta based? Also what was the noise level of each cooler when the delta temps were taken? All of the above info I don't see in OC3D testing is given in Anandtech testing data.