It comes down to testing, one must build a circuit and test said capacitors to find out if there is any truth in statements made by others. There are many factors that are going to change the sound especially in places like filters where the value of the capacitor will alter the cut off point.
Capacitors change their value as they get older which is also going to affect the final sound and changing the original design goals. Replacing older capacitors with new of the same value is going to bring a design back to it's original specifications and the sound will be different to those older capacitors. So, if you have lived with say a phono amp for 5 years and you know the sound of it well and then re-cap it, you will more than likely have a different sounding phono stage, as all the RIAA filters will now be different due to the value of the caps being to spec again.
Now, lets say i have brand A capacitor of value of 1uf 63v that is forming some kind of filter and we then replace that with brand B 1uf 63v. If all specs are the same, then the filter cutoff should stay the same. However, if we didn't measure both of the capacitors prior then the values may vary and the cutoff would then be different resulting in a slightly different sound. If we don't match channels properly as well we then have more variation in design.
So everything must be measured accurately before we can determine if there is any real difference.
To properly A/B we need switchable circuits with matched gain, so we can quickly move from brand a,b,c,d and determine if there is any difference. It all comes down to experimentation.
It's all very interesting if there is any difference, yet it should be done with at least some basis to A/B at a minimum.