This is an interesting topic, thanks to Craig for bringing it up. Back in the beginning days of Axiom, Sound Canada (which later became Sound&Vision Canada) did their speaker testing at the NRC and it included a full set of anechoic measurements and double-blind listen testing. As it turned out this was the peak for reviews based in science.
The reality is that speaker testing involves a lot of expensive gear and a lot of time. To do a full set of amplitude response curves requires an anechoic chamber and enough sweeps to get a total radiated sound power curve, that is expensive and time consuming and you are really just getting started. To do proper double-blind listen tests you will either need a shuffler, like Harmon has, or you need to repeat all the tests with each product in each position to weed-out the position error from the results. Then there is the testing to find out where the upper limits of the output exist and what happens when you reach them. Since this sort of testing involves destroying a few speakers along the way it is quite impractical to do as part of a review. Given all this it is not surprising that most “professional reviews” are really just “professional opinions”, and in many cases just “opinions”
.
