No resistor, capacitor, or inductor is perfect. Just for fun, used an LCR to measure capacitance of a resistor. Understand it is parasitic capacitance, frequency dependent. Sand cast resistor - 1272 uF, Audio grade precision wire wound - 909 uF, Metal Film resistor - 449 uF, and Precision carbon film - 306 uF. Those values seem unreasonably high. For the moment assume the spread from high to low is correct, 4X.
Those that dismiss better quality capacitors in a loudspeaker crossover attribute any measurable or audible change to tolerance and ESR moving the crossover frequency. Assuming replaced parts are not defective, those two parameters could vary a few tenths of a uF and tenths of an ohm in a midrange or tweeter crossover circuit. That amount of variation makes a tiny difference in XSim. Hardly enough to shift a crossover point enough to be audible.
At the same time the capacitor dissing crowd deems sand cast resistors as perfectly fine and better quality is snake oil. Am I missing something?