The actual method of testing isn't as important as including the attenuation due to atmospheric absorption loss. This loss is about 1 dB/1000 feet for a pure tone at 500 Hz and is proportional to the frequency. This, of course, is in addition to the inverse square law loss of 6 dB/doubling the distance. This would give us something real to work with. Of course there are other variables that can't be controlled, but this would at least give us a way to quantitatively compare the performance of one unit against that of another. As things currently stand, comparison is only on a qualitative basis.
We need to stop using the estimated loss of 10 dB/doubling the distance. The laws of physics and acoustics simply don't support it. According to this method, frequency doesn't even enter into the equation, when in fact it is one of the most important variables.