If the 48V is used for phantom power, and ultimately connects to an XLR plug, that may explain some of the weirdness. Phantom power is strange in that the return current for the out-bound device flows through the XLR shield, which is tied to the chassis as it enters whatever enclosure it is connected with. So some consideration needs to be made between analog 0V and chassis, which ideally should just be tied at a single point with no intermingling currents.
I think in an ideal world, there would be 4 conductors flowing down an XLR cable, the two differential audio signals, and a power and ground rail. Pin 1 should only ever be used to extend the chassis along the cable, and never used for anything related to the audio signal. In such a system, the 500 series PCB wouldn't need to consider anything to do with the chassis or it's potential, as it would be all handled by the rack enclosure itself (and the PSU would contain the single analog 0V to chassis connection).
Also in an ideal world, that resistor can be removed and the system should still fully function, with the exception of the weird shield/chassis/PSU 0V mixing that happens in any phantom powered system.