As @snilf states above, "Science gets better and better at identifying, describing and quantifying what is universally subjective, and so, in principle the audible differences between interconnects must be 'measurable,' even if not yet, if they exist at all. But those still hypothetical measurements no more guarantee an agreement in preference than would a comparative chemical analysis of Chateau Mouton-Rothchild and Chateau Lafite-Rothchild."
I agree that most of the differences we hear between cables should be measurable at some point in the future, but we are not there yet. At present, we can measure some things about a cable (e.g., frequency response, resistance, capacitance, etc.), but we don't usually have a clear handle on whether anything we measured actually accounts for differences we hear between specific cables. And even when we hear a difference, it's not a given that there will be agreement about which cable is preferred.
If you haven't actually read the AES publication I mentioned earlier in this thread, you might still want to take a look at it. It's a paper by Milind Kunchur from the U. of South Carolina entitled "Cable pathways between audio components can affect perceived sound quality." It was published in the Journal of the Audio Engineering Society in June. The study included double-blind listening trials comparing two interconnects: a balanced Straight-Wire Virtuoso and a single-ended Monster-Cable Interlink 400. After identifying significant methodological problems with the traditional "short-segment comparisons" of the ABX paradigm, the author developed and implemented an "extended multiple pass" paradigm that is actually a lot closer to what a person experiences when listening to music. (Please read the paper for more details.)
A group of 18 college students participated in the study, and these students completed a total of 59 double-blind listening trials. 43 of these 59 trials were judged correctly; the probability of this result being due to chance was p<.0005. When frequency response was measured, both cables varied by less than +/- 0.005 dB over the range of 16 Hz to 22 kHz. Resistive losses were too small to be considered important. However, the noise levels measured in the StraightWire cable were significantly lower than in the Monster cable.
The author concludes that two system configurations differing only by the interconnect pathway are audibly discernible, even by average listeners with no special experience in music or audio. The study did not complete an exhaustive exploration of all possible factors that might contribute to sonic differences between interconnects. However, electrical measurements did suggest that noise might be one factor affecting sonic performance. The author also noted, "The measurements also show that characteristics such as resistance and frequency response, that naive consumers may focus on, are irrelevant for distinguishing HEA ("High End Audio") interconnect cables."