I still do keep hope alive of someday getting off this thread, but (very simplistically): If you're sitting in one spot, you might react positively to a cord substitution that subjectively tightens the bass response. If you're sitting in another position, even with the same system, you might instead react more positively to a cord substitution that subjectively mellows the treble, or whatever. (The same would also apply to a change in speaker placement, choice of music, etc.) Remember, this is always in relation to the use of some other power cord; it's not possible to compare against no cord at all.* The same goes for listening position -- there is no neutral reference, only preference for a given set of circumstances, and when circumstance changes so may preference.
*[By contrast, it is often possible to evaluate interconnects against the neutral reference of their absence from the chain, rather than purely against one another. It's called bypass-testing, and it's more objective than conventional, subjective substitution testing, but few seem to do it, or even know about it. One easy way to do this, which I use, is by placing two competing interconnects under test into a remote-controllable preamp's A and B tape loops. Then from the listening seat it's a simple matter to, with no switchover delay, select and deselect first loop A a few times, then do the same with loop B, by which method the addition of each interconnect into the chain is compared against the direct feed (the 'bypass', or that interconnect's absence from the chain -- you are NOT switching directly from loop A to loop B; you're switching each loop in and out of the chain in turn, always listening next to the direct feed as your baseline reference). Whichever of the two interconnects alters the sound the least (or least objectionably) as compared with the direct feed wins the round (and can advance to a subsequent round if more than two interconnects are being tested). If you've never done something like this before, you might be surprised at actually 'hearing' your interconnects in isolation for the first time.]
And since I'm posting, I'll risk adding that I don't subscribe to the notion, advanced above, that changing out a whole suite of 'em in unison is the best way to evaluate power cords. The choice shouldn't be between exchanging two different cords between two components, which is obviously confounding, or else going to that other extreme. As with any test, it's always best to hold the variables down to a single controlled one if possible. In this case that means substituting one cord on one component while the others are held constant. In the case of changing out all the cords together, to me that's unecessarily introducing too many variables at once; you may certainly get a feel for the 'sound' of that new model of cord vs. whatever suite of cords was replaced, but not for how that cord synergizes with each individual component, which means you may overshoot the mark and miss the optimum configuration. Of course one may work one's way toward progressively replacing them all with the new cord, after having thoroughly evaluated that model by serially substituting it throughout the existing system (as ultimately happened with me), but I expect that for most audiophiles, like for myself, acquiring multiple duplicate cords before having even comprehensively evaluated a single example represents a putting of the cart before the horse and is unrealistic. (I also don't buy the notion, however helpful and comforting it may be, that one must hear any changes obviously and immediately or else they are rendered definitionally insignificant.)
*[By contrast, it is often possible to evaluate interconnects against the neutral reference of their absence from the chain, rather than purely against one another. It's called bypass-testing, and it's more objective than conventional, subjective substitution testing, but few seem to do it, or even know about it. One easy way to do this, which I use, is by placing two competing interconnects under test into a remote-controllable preamp's A and B tape loops. Then from the listening seat it's a simple matter to, with no switchover delay, select and deselect first loop A a few times, then do the same with loop B, by which method the addition of each interconnect into the chain is compared against the direct feed (the 'bypass', or that interconnect's absence from the chain -- you are NOT switching directly from loop A to loop B; you're switching each loop in and out of the chain in turn, always listening next to the direct feed as your baseline reference). Whichever of the two interconnects alters the sound the least (or least objectionably) as compared with the direct feed wins the round (and can advance to a subsequent round if more than two interconnects are being tested). If you've never done something like this before, you might be surprised at actually 'hearing' your interconnects in isolation for the first time.]
And since I'm posting, I'll risk adding that I don't subscribe to the notion, advanced above, that changing out a whole suite of 'em in unison is the best way to evaluate power cords. The choice shouldn't be between exchanging two different cords between two components, which is obviously confounding, or else going to that other extreme. As with any test, it's always best to hold the variables down to a single controlled one if possible. In this case that means substituting one cord on one component while the others are held constant. In the case of changing out all the cords together, to me that's unecessarily introducing too many variables at once; you may certainly get a feel for the 'sound' of that new model of cord vs. whatever suite of cords was replaced, but not for how that cord synergizes with each individual component, which means you may overshoot the mark and miss the optimum configuration. Of course one may work one's way toward progressively replacing them all with the new cord, after having thoroughly evaluated that model by serially substituting it throughout the existing system (as ultimately happened with me), but I expect that for most audiophiles, like for myself, acquiring multiple duplicate cords before having even comprehensively evaluated a single example represents a putting of the cart before the horse and is unrealistic. (I also don't buy the notion, however helpful and comforting it may be, that one must hear any changes obviously and immediately or else they are rendered definitionally insignificant.)