Okay, incoming coffee-fueled saturday morning write-up.
Prescriptive targets are the amount of gain recommended at different levels and frequencies for a particular hearing loss. NL2 (Australia) and DSL (Canada) are the modern winners for adults (although people will still use NL1 in certain places, or even CAM) and they have converged over time as research has tried to optimize them, which you might expect they would if there is truly some sort of platonic ideal for gain prescription. There are still differences stemming from underlying ideology of how they were developed and you can generally expect NL2 to be a little bit more compressed, less loud/sharp. Currently research comparing user outcomes shows that people do well in both prescriptions with DSL maybe having a very small edge in noise situations, but the evidence base there is too small to really lean on. For children, DSL has separate child targets that are generally the only ones used; they have a strong focus on audibility are and significantly louder than adult targets. (Note, users who grow up with child targets are not and should not be transitioned to adult targets once they are adults. Adults with later onset hearing loss may do very well in child targets, but many cannot accept them due to loudness tolerance issues). Proprietary targets are developed by manufacturers largely based on NL2 and reasons for their differences are mysterious and may include increased initial user acceptance (# hearing aids sold), marketting (“our hearing aids sound ‘different’”), and internal research that suggests some change is good either based on small-n internal research or more recently large-n datamining (yay!). Historically, proprietary targets have on average just been quieter than independent targets which would certainly be the case if they were focussed on intial user acceptance. I looked closely at proprietary target implementation of current hearing aids last summer when I was discussing targets with Volusiano and that appeared to still be the case on average with the exception of Oticon VAC+ which is applying more gain than NL2 at 2kHz.
Whether a hearing aid is actually delivering prescriptive gain to your eardrum depends on the acoustic coupling and shape of your ear canal (effects can be huge), as well as the implementation by the manufacturer. There was research a while back showing that what the manufacturers called “NL2” actually differed quite dramatically from actual NL2 targets and from one manufacturer to another. Things have gotten better since that was ‘outted’, but there are still differences. DSL is a little bit more likely to be implemented in a static way across manufacturers because, as far as I recall, manufacturers need permission to use DSL and have to demonstrate to the National Centre for Audiology (that developed the targets) that they are implementing it following their guidelines. So the job of REM is to ensure that you are actually getting prescriptive gain and, yes, if you set a hearing aid to a proprietary strategy and then verify it against an independent stretegy and wrangle it to those targets across levels then you are essentially just turning it into the independent strategy. But remember that audiology as a whole is sort of a game of fuzzy precision. If you were fit to proprietary targets and then complain that the hearing aids are too sharp and the clinician rolls off the highs, you are now no longer using proprietary targets. If you do not have a perfectly kemar-average ear and first fit a set of hearing aids, you are not using proprietary targets. If the clinician only adjusts frequency-specific gain but not level-specific gain using REM, then compression strategies will still be proprietary although frequency strategies are now independent. And then the whole game is complicated further by neural adaptation and the limits thereof (e.g. hearing aids can sound one way at first and then your brain just readjusts over time and they sound natural again, so how much should you be tweaking them to comfort versus just waiting for your brain to get there).
You might imagine that if you set a hearing aid to NL2 and adjust it with REM so that it is taking into account your individual ear and meeting targets appropriately, then when you switch to DSL targets the hearing aid will just maintain the gain changes that were done to meet REM targets and apply them to the new prescription. This would be lovely, but it is largely not the case. There are even differences between manufacturers on whether they will do this for changes in audiograms. For example, if you input a new audiogram into a Sonova hearing aid it will (try to) simply adjust for audiometric changes and maintain fine tuning, but if you input a new hearing test into a Demant product it will try to just first-fit it again (annoying). For this reason, a clinician may regularly update the audiogram in the first product, whereas they may never update the audiogram in the second product for the life of the device and will simply readjust with REM for audiometric changes because they don’t want to totally refit.
Inputting a well-measured RECD into the software, if the software allows it, will typically get it closer to targets at first fit. The effectiveness of in situ to try to address this is, in my experience, not great. It’s maybe(?) better than nothing in a situation where a clinician does not have REM, but it is not generally the intention of the manufacturer that it be used in conjunction with REM in places where that is available–keep in mind that manufacturers are also trying to develop products for international situations where not only is REM not available, but audiometers might not even be easily available.