You misunderstood my post. Let me explain again: the goal with battery usage is to get rid of SMPS leakage current, and any capacitive coupling to the power grid (and earth/ground connections, of course).
I understood you perfectly fine, I was just wondering if the usage of a powerbank wouldn't introduce other issues that might not be obvious immediately.
Anyway, I noticed I never confirmed it explicitly, but you were right, of course: the noise in my setup is indeed not audible when listening through the headphones, so it is indeed the active monitors by themselves that make the noise audible.
Today I received a Radial Ice Cube IC-1 balanced line isolator:
https://www.radialeng.com/product/icecube
Plugging it into the XLR chain from ADI-2 DAC to active monitor first did nothing, noise was unchanged, but engaging the ground lift on the IC-1 made the noise go away.
EDIT:
Doing a few more tests, and changing the position of the active monitors in the powerstrip, the noise actually came back, even with the line isolator and ground lift engaged, so that points to the power connection of the active monitor that's picking up the noise. I'll be attaching an UPS to try and isolate the issue.