Whether it's "too little to be audible" depends on how sensitive your speaker driver is and what the noise floor of the room is.

In the video I linked, he was unable to induce any noise into the speaker wires greater than -130db at 60hz. The threshold for human hearing in absolutely ideal circumstances is about -115db. Anything less than that is provably inaudible for all frequencies, and for 60hz the threshold is much higher than -115db. The noise floor of the room only raises that further. For perspective, the very best state of the art DACs can only keep noise levels to -123db and just a few years ago the number was significantly worse.

The design of an amplifier stage often looks like this:

https://toshiba.semicon-storage.com/us/semiconductor/knowled...

Where the triangle isn't perfectly linear. If any RF gets into the output, it can be converted to audio frequencies by the nonlinearity and amplified. The open loop gain of the amplifier component can increase the magnitude of signals like this by tens of decibels.

If noise gets into the pre-amplified signal then it won't matter which wire you use to connect the speakers. This is why ground loops are bad.

> If noise gets into the pre-amplified signal

Nah-- it's about high frequency noise getting in via the speaker lead, and coupling to the negative feedback of the amplifier stage, and getting rectified and amplified.

Yes, most interference is via line level inputs, but it can easily be via the "output" as well.

And yet in real consumer hardware this is never an issue. Do you just like wasting people’s time?