Mosdef.
Worthless is an over exaggeration, oops.
Not accurate, as impedance varies with frequency. Obviously to precisely set gains, you would need an oscilloscope and a non-reactive load.
My point is, other than matching levels (which I would do unloaded), I find setting gains with a DMM to be a waste of time. A trained ear works much better, with a variety of music, and final adjustment via the HU.
The idea that setting via a DMM is the magic way to correctly set the gains on your amplifier is ignorant.
using the word "magic" in your post is passive aggressive. Ohm's Law is not magic. electrons behaving as particles AND waves is close to magic.
I agree with you from a technical standpoint, but not a practical one. You can't ask everybody to use their ear - many people think clipping is what music sounds like loud, and the music selection is crucial - nowadays people listen to compressed music that is clipped in the recording.
the test tone used for DMM gain setting should be sufficiently higher than the fs of the driver, so the sub impedance is at a nominal level. asking people to break out their fully-rated non-inductive dummy load is ignorant - no one has one. so let's be realistic in our recommendations by recognizing what tools average people have on hand. a DMM is a standard tool that any installer should have. if someone doesn't have a DMM or if doesn't know how to use one - they should not be installing car audio. a basic understanding of Ohm's law and basic DC electricity is important. but very few people in any car audio forum have access to an oscilloscope #1 nor an understanding of how to use it #2. This is why SMD developed the DD-1.
using an oscilloscope or using a trained ear are not reasonable recommendations to a general audience for any car audio forum.
using a 0dB test tone at frequencies that are far enough away from driver fs is a fine method of determining where gain (input sensitivity) should be set to obtain a specific power output. Ohm's law works well here. the real trick is knowing what power output to shoot for since amplifier rating methods are so inconsistent. yes, music levels vary - but we don't care about levels under 0dB recording - just about what a worse case signal level is. most modern music is mixed using an "equal-loudness" technique which results in most of the recording being close to, or at 0dB. using 0dB test tones is the best way to ensure that the test is representing worse case conditions - to prevent clipping due to signal amplitude. from there, you can account in signal recording strength in your use of the volume knob.
another challenge (regardless of the method used) is determining what HU volume level to use when setting gain. but this is a different topic as it is independent of measurement method.
your method by ear and using a variety of music is risky to ask of anyone. how do you know the music is clean (i.e. not clipped in the recording) and that it represents a true worse-case signal level at frequencies of interest? with software and an understanding of recording methods and signal chain gain structure someone can choose test recordings - but we cannot ask that of everyone. Autosound2000 discs have test tones with a few seconds of clipping to help the listener identify clipping and ensure it is not present.
we are here to K.I.S.S. and give recommendations that are feasible for the intended audience.
engineers usually struggle with providing reasonable recommendations that can be implemented by the average person.