Hello TiltedTowers,
I think that the easiest way to solve such task would be to create a new chunk each time a new multi-camera system should be added, then make the new chunk active and add the images as multi-camera system to it. If necessary - repeat several times for all different multi-camera systems used. Then merge the chunks, set merged chunk as active and remove the others.
Such approach would keep separate calibration groups for multi-camera systems from different chunks (so the sensors would not be merged with the chunks, they will remain as separate instances).