That is indeed true but it does not mean that a decent quality trained ear can’t hear the differences in resistor qualities in a well executed circuit. so the idea of swamping does not hold true.
Apples and oranges. As stated previously everything has a sonic signature and everything interacts. Agreed that not all can hear signatures.
Unless one matches capacitors or matches resistors to measured C, there are audible deltas that some may misattribute to component rather than FR errors. A single 5% part at its ± value extremes yields ≈1.9 and 2.1. Such deltas are readily audible in the critical 2-5kHz region.
Some manufacturers go to such lengths to match values so the product matches the spec. One was Spica. When I rebuilt my TC-50 XO, I measured the drivers, inductors and resistors and then selected capacitors to match specification as closely as possible. The imaging improvement was astounding relative to the aged originals, all of which still measured within their tolerance. We could debate the contribution of replacing EL & Mylar w PP vs. tolerance compensation.
Bottom line: E V E R Y T H I N G matters