It's a surprisingly-complex domain!
Simply modelling all of the relationships is a fun experiment.
i.e. so-and-so is the producer and advertised co-artist, so-and-so is both an individual artist but also a member of bands X and Y, bands X and Y are part of the super-group Z; label A is a subsidiary of B, which was a subsidiary of label C from 2012-2016, was independent from 2016-2017 and is a subsidiary of label D from then until now.
Now, think about building RBAC on top of this, like you're hinting at!
No small feat, to be sure.
edit:
The main rabbit-hole I dove down was revenue-sharing, so that labels and artists can be paid out immediately, without the latter needing to wait for some settlement layer governed by the label. It turns out Stripe allows you to implement this fairly easily, and with limited fuss!
Even if the label chooses not to use this process for some reason (I wonder why!), you should be able to model out revenues for all parties, and show the artist(s)/song-writer(s)/label(s) what they should have had in gross and post-tax earnings for any given period in time.
You can even show the music consumer what % of each purchase will go to the various parties involved in producing the music, which might help in steering people towards indie labels or artists. Kind of like nutrition facts for music. :)
