I have no beliefs that AC cable upgrades make any audible improvement.
My speaker cables are all generic 12 AWG. I've read enough blind cable comparison tests & measurement results over the years to feel confident that I'm not missing out on even a tidbit of sonic excellence by not upgrading to some esoteric $1000 per foot speaker or AC cables.
May I commend your use of the word "believe" -- because, much of the time, it is as much a matter of belief as it is of "acceptance" of scientific evidence (such as it may be).
(Caveat: further down I mention measurements which I do not have, unfortunately. For anyone reading on, you can take this as being anecdotal!)
As it happens, I was incredulous about AC cables and set out to try the after-market variety after having chatted with an engineer who sweared by his own "AC chord recipe". (Note that my amp comes with "designer" AC chords anyway Symphonic Line's own "reference" cables).
This meant that I purchased good quality connectors and about 25m OFC wire (single strand). I crafted the AC cable as per recipe, checked it, and connected it to the DAC and set out to play some music.
I immediately perceived a difference (more clarity is what I remember), more so after about 10 mins. It was easily discernible, blind or otherwise.
I found this pleasantly strage; pleasant, because I like the sound, and strange -- for obvious reasons. I conjectured that, if there actually
was a difference, it would show up on a FR curve -- right? Because the difference I perceived (or thought I perceived) was in the sound...
So I visited a friend who has a full diy set-up with B&K mic and horn-loaded speakers with Murata super tweet, etc, the works. We decided to use my cable on his amp ( 2A3 based tube).
To avoid the possible effects of switching on & off, we let his system play with each AC cable before using pink noise, & some music (Pink Floyd was one). When it came to the actual measurement, we used one speaker, then both speakers ( don;t remember the reasons for this).
Whaddya know, the response curve was slightly different, in the upper mids & high frequencies as well. We measured at high amplitudes.
Unfortunately I do not have the graphs as mentioned above.
We casually attributed the differences to better connectors, single-strand wire...
For what it's worth, better connectors should result in a better mechanical contact, so perhaps there's a reason there. Then there is the single strand rather than the usual multistrand (which is more flexible btw)...
Needless to say, the pleasure of this experiment was enahnced by a bottle (or two) of red wine.