I don't think there's any rigorous method that can be applied (as you usually can with substitutions when tracking down a specific wiring problem).
You could serially get a much more expensive/highly regarded model of each component in the system, and swop it in, and see which substitution produced the biggest positive change, but then you'd be ignoring system synergies to say nothing of issues of output and input impedances. And other variables as well.
Most of us, most of the time, go on gut feeling. Reinforced, perhaps, if one of the components is older than the others or not so highly regarded or notionally "out of line" of whatever you think that component should cost as a % of the total system.