Driver, I'm aware that many companies short the two pins and it may be acceptable in some cases - but not all.
Scenario#1: input stage is made of bipolar transistors in a balanced differential amplifier circuit. In this case, shorting the input actually changes the bias (DC) current through the transistor base. The circuit may seem like it's still "working", given that one half is still biased correctly, but that's not what the designer tried to do. It may in fact distort the input signal in this configuration.
Scenario#2: Input circuit is not DC-coupled but it is differential. In that case, shorting one input makes the circuit look into a zero impedance for the AC signal, not the Dc component. Some circuits may not operate optimaly under these condition, which are in fact an imbalance in what's supposed to be a fully-balanced circuit. This may have an effect on the distortion components at the output of the circuit, even though there is no signal at all at the shorted pin. The issue is that the balanced circuit "reflects" the signal to the negative side by virtue of modulating the voltage at the joint node.
I hope this gives some idea why shorting is not ideal. On the other hand, a low resistance would not have much of an effect on noise in the circuit and it would avoid the complete short. The ideal solution would balance the load on both ends for AC and DC, but that's a tweaker's solution, not a practical concept for incorporating into a cable...
Scenario#1: input stage is made of bipolar transistors in a balanced differential amplifier circuit. In this case, shorting the input actually changes the bias (DC) current through the transistor base. The circuit may seem like it's still "working", given that one half is still biased correctly, but that's not what the designer tried to do. It may in fact distort the input signal in this configuration.
Scenario#2: Input circuit is not DC-coupled but it is differential. In that case, shorting one input makes the circuit look into a zero impedance for the AC signal, not the Dc component. Some circuits may not operate optimaly under these condition, which are in fact an imbalance in what's supposed to be a fully-balanced circuit. This may have an effect on the distortion components at the output of the circuit, even though there is no signal at all at the shorted pin. The issue is that the balanced circuit "reflects" the signal to the negative side by virtue of modulating the voltage at the joint node.
I hope this gives some idea why shorting is not ideal. On the other hand, a low resistance would not have much of an effect on noise in the circuit and it would avoid the complete short. The ideal solution would balance the load on both ends for AC and DC, but that's a tweaker's solution, not a practical concept for incorporating into a cable...