Larry,
I have not done listening tests with 15m interconnects.
I have done impedance checks with 10m + cables and noted the increase in Zout is merely the addition of the resistance of the wire and the Zout of the GK-1.
All the physics points to a potentially strong influence on the sonics from some interconnects of this length. But with the tube topology used, I am aware from my reading and anecdotal information that this output configuration will have less susceptibility to adverse sonics than any other.
You need to decide if this is an outrageous claim. In the absence of conclusive empirical evidence it is difficult to say until you do it. Accordingly, I have qualified my comments, but because I have not done listening tests with 15m interconnects I cannot promise predicable results with all systems. Larry, think about this - how on earth could I, when people report significant differences with 1.5m intereconnects? I've done most things with audio, but not all of it......! (more's the pity, I'm working on it!)
However, let's be constructive and solve the problem. Make a 15m length of twisted pair CAT5 cable, and try it out. Keep the wire linear, not coiled. It will cost you about $AUD10, and will conclusively demonstrate in your system whether or not the concept is viable. If the effects are minimal, then extend the concept to more expensive material to lessen the effect further. If it were me, I'd be acquiring some satellite dish coax and cross connecting it as in Thorsten Loesch's configuration; this is fairly cheap, a lot lower than $25 a metre, and should give excellent results.
I hope this is helpful, but I am unable to give assurances that unconditional addition of a 15m interconnect on the GK-1 will go well with any system. That would be neither prudent nor honest!
Cheers,
Hugh