these voltages really is not significant to the solution to your problem but it is obvious that it is not the applied to the brushes voltage. That should be about 14.15v.

I suspect that the actual test would be to set up the supply and measurement point then rotate the armature through all brush pairs. The actual reading would not matter only the acceptable range of the readings need be known to determine if the rotor has an internal failure. This is based on the fact that the winding pattern used is continuous. With enough data sets you could also determine the winding pattern but that seems unneeded as the brush setup and wiring on the motor at disassembly should give you that data.

We are back to the original question of measuring a voltage and ensuring that the amplifier inputs remain in the acceptable common mode range. That no longer seems difficult given that the supply is now 25 v maximum and the currents are buried in the windings.
Send the collected data through a(digital thus easily set) window comparator, flag the out of bounds readings, trip an alarm, print them in red on the print out or do what ever you need to do with the data.
Terry