If you test at half octaves, you double the representation of the loss. Most REM’s sample at about 100 points across the spectrum too. The general theory is to reproduce the proportionality of the long term speech weighted average kayak, which is neither a line or always curved when re-produced in term’s of the dB SPL missing relative to somebody’s loss. (Looks nothing like my kayaks either TBH…)
Most systems provide the opportunity to include patient loudness growth data, but seeing that is massively subjective and not hugely helpful, it tends to be a step too far.
As the good Doctor says above, valuable for ratification and problem solving in unusual cases, though there is a methodology that says if you REM everybody, you cut down the number of subsequent adjusts (M. Poe’s Law).
What’s the issue with a good Audiogram? it’s not that difficult to get one with a decent set of inserts and a quiet room: I acknowledge the subjectivity, but the more than doubling of power over each 5dB tested means that it’s almost impossible to be more than one measurement out if the patient understands the test and isn’t appreciably wobbly in their own interpretation/has tinnitus.