I don't want to get into a passing match about cables but they can change the sound. Cables are transferring an electronic signal to a dynamically responsive device that can and will change its frequency somewhat based on the signal being sent through it.
At the very least we can agree that amp impedance changes the way a pair of headphones sound then why is it so offensive to say that cables that dirty same thing but on a smaller scale do the same exact thing? It changes the signal that the headphone receives and therefore changes the frequency response shape.
Another proof is to look at innerfedility graphs and see how the impedance changes as you increase frequency range. The only exception to this is orthodontic which tends to be equal from bottom to top.
The real question and debate I think you're trying to make is whether it not its worth it? That is up to you and a matter if opinion and not a means of dismissing proof on paper that cables can change the way something sounds. As IEM's and headphones get more and more revealing and transparent it's easier to tell.
Personally I think there's a line and that's beyond the comfort and aesthetics that makes it not worth the point of diminishing returns.
not saying a cable cannot change sound compared to another cable. I've been calculating an impedance change in the cable just last page to explain the sensitivity change from p to s. I just need to do it for 2 frequencies while looking at the er4 impedance curve instead of taking the 1khz spec to get the actual FR change at those frequencies. in fact I actually made a very poor excel file to avoid having to redo the calculations of each frequencies.
you can see an example where little impedance change can be clearly audible in that review at the bottom http://www.head-fi.org/products/fitear-to-go-334/reviews/10484
so of course I think impedance changes in a cable+source can change the sound of at least some IEMs.
what I'm saying is that an average cable is already low impedance, because there is no reason why copper wouldn't conduct well, if not then the cable is bad and that's it. so getting a better cable that goes down from there like using silver, won't get us to -10ohm. it's only moving from maybe around 1ohm to a little less, and that by itself has little chances to create a significant change in signature.
and the er4 doesn't go low enough in impedance to be one where it really matters. and of course as I said, with the er4s(so with the added 75ohm, changing cable wouldn't mean shiit when it comes to impedance in regard to what has already been added.
now the se846 is a counter example and one I have been using to tell people not to buy a pono or a sony ^_^. but that's because added to the wide change in impedance, that IEM goes super super low, golden ears gives it going as low as 4ohm
. so obviously just 0.5ohm change from the source could go close to 1db change(rule of thumb but I could make an actual calculation). for the er4 0.5ohm would get a much lower change, one I would call insignificant to my ears.
now some cable change more than just impedance, after all a cable plays a perfectly known game with resistance, capacitance, inductance, and freqs. there are ways to mess around with those and get an audible change, but actually little ways to improve the overall signal fidelity because a basic cable is already pretty great at doing the job.
stuff like crosstalk can be improved if the cable runs 4 wires all the way down to the plug, same if they are well separated and insulated, or it could be worst if the shielding is massive...
plenty of ways to make a change, little ways to make it audible, and even less ways to make it an actual improvement(as in what came in goes out without change).
so yeah a low impedance source is the way to go on most IEMs, and a cable should be bought mostly for comfort or how pretty it is, or if the original cable is real known crap. is what I believe.