Fidelity can be defined as signal transfer accuracy.
As sound is definitely a signal, we can speak about sound transfer accuracy. "Quality" on the other hand is where the trouble lies. What is the definition of quality? I'd say the definition of quality is subjective. So if we want to talk scientifically about sound quality, we should be actually talking about sound transfer accuracy (fidelity). Or we can dwelve into philosophy or psychology. Which might or migth not be a suitable topic in "Sound science". Depending on the subjective definition of "Sound science", of course

Tyll stated, that the recordings of Django Reinhart have lousy fidelity. I totally agree. But these recordings are nowadays the only source of his music. We have to decide what is the best way of reproducing these records. Most of us would probably say that the best way is to reproduce the signal on those records as accurately as possible. But can we perhaps alter the signal so that the perceived sound would match the man with the guitar live, lets say year 1930? Possibly.
I think my point is that the definition of sound quality is too vague to really account for a meaningful discussion.