well…it’s the spec everyone uses in the music industry.If you try to do a root mean square average on power, you get the wrong answer. Root mean square averages only work on parameters that need to be squared to give power.
In text books, one usually find the explanation of the root mean square average alongside a sinewave, with a derivation of the DC voltage that will result in the same dissipation in a resistor. The association between rms and a sinewave, has somewhere been misconstrued that rms is a property of sinewaves. So "rms power" has be a misnomer for sinewave power i.e the power dissipated in a certain load resistors by the largest amplitude the amplifier can produce at a given level of distortion. In order to make the figures better, use a higher level of distortion.