Further to David's comments, a hidden advantage of the old school impedance-matched pro audio method is that not much RF is going to get through the lousy input and output transformers, even if common mode cancellation due to balanced wiring might fail beyond audio frequency. Repeat coils were often inserted into lines for additional isolation. There will be some random coupling due to stray capacitance and whatnot, but there always is.
The goal was rejection of hum and LF common mode interference more than RF, although twisted pair should offer some immunity to RFI.
As far as I am aware, no audio cables of the era had controlled impedance, although you can buy twisted shield pair cable today that is standardized (for ethernet and whatnot) 100 ohm being the common impedance for twisted pair. But these work at HF frequencies.
This notion of RF in high feedback solid state amps was a hot topic in the 80s, but I don't recall seeing any earlier examinations of the problem or studies involving tube amps.
Reflections in speaker cable seems a big problem if you want to introduce it as a variable. First, where is an 8 ohm speaker 8 ohms? Secondly, a transmission line model assumes that voltage and phase will vary at different points of the cable, it is an impedance transformer. Thirdly, dedicated engineering is required to come up with an amp which will have an output Z matching speaker impedances....even zero feedback output stages will only get you up to a couple ohms.
Re #2: Who needs a speaker cable that you can't cut to length? Don't we have enough problems already?
Today's experiments with current amplifiers having a high output impedance, equal to or even higher than the speaker impedance, head in the direction of impedance matching. This might be a good approach but it throws away the high damping factor "load independent amp" scheme that has been so dear to audio engineering since the 1950s.
I got to sit through two experiments where amps were switched from voltage mode to current mode via variable current feedback schemes, one with jc and Bae at Silbatone and another by Menno van der Veen at ETF. In both cases, I greatly preferred the current drive scheme, way more psychedelic and organic, but at ETF, Jan Didden, editor of
Linear Audio greatly preferred the tighter high damping factor voltage scheme.
Damn Dutch Protestants! I married one and I still can't figure them out!