mains cable won't make any difference
The snake oil exists about Mains cable I would agree with, actually any cable.. for me it's often a compound problem - capacitive loading, inductive loading (both the cable adds), impedance, shielding and in the case of RF the material.. finally how much current can you get through it. The rest is typically currents caused by ground or really bad routing.. that then couples and causes problems.
Mains - ignore the cable but treat the noise.
I always use an IEC filter on my mods and designs. It may not be much of a change but that change is made before the noise gets into the chassis of the audio gear. You can use a SMPS and hit -160dB.. no noise.. but you need to understand that and not think a simple power supply filter will do the job.. voltage regulators (including the low drop out low noise) noise rejection tops off at a max of 1MHz.. after that you're onto ferrites like mini chokes (even surface mount ones) and getting the right board layout so power being sucked into a chip doesn't then make a massive amount of noise across the entire system.
Proper professional mains conditioning systems (like all the bands use when they tour) take in whatever.. and output very very load noise (and often remodulated 50Hz) power. No matter how bad the local power is.. what goes into the amps and audio systems is cleaner than most audiophiles power.
50% of an amp's ability is it's power supply. 49% is the amp design and then 1% is components.. if you have bad power you have a bad day. If you have a bad amp.. then you have the ability to fix it. Then that last 1% is selection and matching of components.
Digital and Audio have different behaviours.
Same signal.. except that DC return current will follow a direct path of least resistance (also most audio being under 20KHz), whereas once you get about 100Khz and definitely by the time you get to 1MHz or above, the return current follows least impedance which is usually *under* the track even if there's a ground plane. Now one you get loops with cables.. that may then be the path of least impedance.. and you have a noise problem.. even if you have a ground plane or a shield around your cable.
To reduce this - isolate. it breaks the loop and allows the high frequency to terminate and return current to flow back.. the isolator then replicates the onward signal and the return current of that is on the other side of the isolator thus no loop.. and low noise..
Ignore snake oil, find out why the cap sounds better.
I've seen some really funny ridiculous claims. Yet it may be that an expensive cap sounds better than a mainstream £1/100 cap does. However the part I refuse to accept is that claims made without evidence (not being able to say why it sounds different due to hard cold measurements). I'm not saying that because it doesn't have all the measurements a cap is snake oil but I want to see what specifically is causing it to be different (and not because it's new old stock.. is the chemicals, is it the design, is it the resulting harmonics or the impedance spectrum.. it's not for me to guess it's for manufacturers to really provide the cold numbers. Otherwise two identical (cold numbers) that are claimed to sound different is simply hookem.