If the impact on frequency response is <0.1db there is little(no) evidence we can detect a difference and even more variance at the upper end of the spectrum to detect a difference. Now try to find a cable not inherently directional, i.e. with circuitry that has 0.1db difference in the audio spectrum by changing direction. Feel free to use square waves for tests with bandwidth limiting and real speakers for transmission line effects.
The next question to ask is what constitute a "good cable"? Who gets to decide? There is a flaw in this logic.