I personally don’t think there’s an additive double-dipping effect per se. The compression in hearing aids is there to compensate for the varying degree of loudness perception due the hearing loss. You can’t hear a sound at softer volumes below a threshold volume due to your hearing loss, but if that sound gets loud enough, you can hear it OK. So you need a certain gain amplification to get past the threshold volume where you recognize that there is a sound, but if that sound gets really loud, that same amount of gain at the softer level can just be too loud for you tolerate at the louder level. So if anything, the compression at the louder input range is an attempt to “linearize” your hearing to help make it more closely resemble the normal (shall we say more linear?) hearing.
If you have light to moderate hearing loss, there’s enough dynamic range in your hearing for compression to help make the sound volumes seem more linear. But with severe to profound loss, the hearing threshold volume is so high that the dynamic ranges get squished to almost nothing, so linearization is much harder to achieve.
Yes, you can reduce compression to below 2 in music program if you can, to help get more dynamic out of the sounds. But depending on your hearing loss, it’s not always possible. The very narrow dynamic range of a severe to profound hearing loss has very little to no room for you fit the expanded (minimized) compression in the first place.
What musicians do to impart their own dynamics to music (whether they try to linearize it or expand it or compress it in their own way) is really a separate and different thing from what HAs try to do with compression to help restore the perception of linear hearing for the patient.