If anything would have been damaged, it would have been the output stage of the Ref5SE, but that seems very unlikely also. And per your comments the Ref5SE works fine with the different balanced interconnect.
But if you are saying that the original balanced interconnect now sounds 6 db lower than the replacement, with the switch in the balanced position in both cases, I don’t know how to explain that. If you are sure that is the case, all I can suggest is that if you have a multimeter check for continuity between each of its three pins at one end and the corresponding pin at the other end. And check that no continuity exists between each of the three pins and the other two (doing that at just one end is sufficient).
Some clarifications to your first paragraph: The reason the input impedance is 200K balanced and 100K unbalanced is simply that the 200K reflects the sum of the input impedances of each of the two balanced signal legs (XLR pins 2 and 3), while the 100K is the input impedance of just the one leg that is used to receive a signal in unbalanced mode. The reason for the 6 db difference in volume you mentioned in that paragraph, if I’m understanding what you wrote correctly, is that in balanced mode the amp responds to the difference in voltage between the two input signal legs, and when one of those legs is grounded (i.e., connected to 0 volts) that difference is only half as much as it would normally be. Finally, re your first sentence, on USA equipment it is most likely pin 3 that is grounded by the switch, not pin 2, and ground would be applied when an unbalanced signal is being provided, not a balanced pair of signals.