Is it still using the twin pump setup?

The ECU calculates fuel pressure based on actual injector on time vs what it should be based on injector offset voltages. It can tell fuel pressure is low if it has to fire the injectors longer than calculated to maintain afrs.

I guess, I'll give an update to this.
Thanks to the thread on DTC logic and error classes.

I did some digging through code to find where and how this DTC is set. Found some function that is unique to RS6, so I can't really cross-reference it with other "defined" software and name vars/consts properly, but apparently it is responsible for controlling the fuel flow. It also can set P1063 and P1020 DTCs. It interfaces with other functions via few RAM vars, and those are mostly used by EKP ("elektrische Kraftstoffpumpe") routines. I did not invest enough time into figuring out the complete logic of this process, but apparently ECU checks back on fuel pump status reported by the fuel controller and limits the load via LDRLMX if something's wrong. It also sets P1065 if the signal is missing completely.

Also, for shits and giggles, I reverted to the 4 bar FPR - and surely enough, the DTC is still there. It also reappears almost instantly on idle after clearing it.
So, after all, dd was right - it seems that it's the "missing fuel controller" that produces this DTC.