Since the feedpoint impedances are all different then the attached ATU would have different settings, and thus, efficiencies, for each attached feedline.
So how does one easily determine the efficiency of just the balanced feedline using simple easily available equipment?
The below approach is clearly incorrect in many aspects. A better approach could be as follows:
At least for balanced feedlines of known characteristic impedance the test procedure could be as follows.
1) Terminate the ATU with a resistor having the same resistance as the Zo of the feedline.
2) For each frequency measure the efficiency of the ATU using a scope with a 10X probe on the terminating resistor. Derive Power = Vp^/2R. Convert power to dBm.
3) Terminate the feedline using the same Zo resistor and measure the Power dissipated in the resistor as per 2 above.
4) Subtract the result of 2 from the result of 3 to derive the feedline loss in dB.
Do the above procedure for each frequency of interest and for each feedline.
NOTES: Using the same scope to measure the Transmitted power and terminated power for each frequency will eliminate any error introduced by the scope since it is a comparative measurement.