What I find interesting is that the second highest rated cable in the test had the worst fidelity, and the worst rated cable was right there in the middle when frequency and amplitude were measured.
As i noted above, we have to distinguish between accuracy and musicality. They are different and valid. The problem is one is harder to measure.
Most of our queues come from music theory. Concepts of consonance and dissonance, chords construction all follow the "some distortions" (e.g.: harmonically related additions) are welcome. Chords are examples, as is the resonance of a sounding board or violin case. Heck, tubes probably get their great rep from consonant harmonic distortions and that is known to be musical from folks who make music. No issue there, as logn as we understand what is happening.
in cables we also have the possibility that what they are is long, expensive, fixed tone controls. We KNOW that rooms are far from flat. We also know that many electro-mechanical components are far from flat (cartridges, speakers...). Finally we know that there are HF distortions from the digital recording mastering process which, when it is done badly, are objectionable. all these problems can be made better or worse through simple frequency manipulation. So a cable that distorts - rolls of the HF - may sound very good to many in a particular system. Sadly it may sound dead in another.
Decades (?) ago Mark Levinson launched a $20k equalizer on the world from Cello and got great reviews from Stereophile among others. I’m convinced we miss a trick by not having a loudness contour - in fact i think its one reason why many audiophiles listen more loudly than they might otherwise. Of course we need to avoid a cure that’s worse than the disease.
So i’m not surprised.