If this is really an avenue you want to pursue, you could mount a 1 spot in the amp and build a 5 pin cable, din or xlr(or run a separate 2 wire zip tied to your switch cable).
But that all seems overkill...
Also, 50/50 on that DC coming from the amp being very clean. Possibly just some zener diodes between a DC rail and ground.
It's funny you mention using the one spot at the amp because I've already been doing that using a TRS cable. Audio on one line and power on the other. It works great.
My next experiment is going to be running them both on the same line, like true phantom power but, using a combiner box at the amp and break out box on the floor. It will be unbalanced, but as long as there is some power filtering during the break out, I think it should work.
I would advise against that. Issue being that if there is any audio freq noise at all on the power rail, it will end up in your signal path. This is solved in pro audio with the fact that the audio is balanced so any noise is phase canceled.
Also, if your decoupling (DC blocking) ever fails or shorts, you could potentially send DC to your guitar, thus send DC to ground. Also, DC to the input of an amplifier(which *should* have decoupling built in, but still...
I'll defer to Dr. Ian Malcolm:
I hear the warnings and I am hesitant to proceed, but I think this is one of those things I'm going to have to test for myself.
I think a 15 ft shielded guitar cable with an instrument level signal should have relatively minimal noise compared to the longer mic level cable runs that a balanced signal was designed for. A high pass and low pass filter on the audio will hopefully take care of any noise.
We do DC blocking caps on the ins and outs of every pedal, so couldn't any pedal potentially fail and send DC to the guitar? Plus, I'm not convinced DC on a passive pickup is an issue. A pickup is basically a 3k-25k resistor. Please correct me if I'm crazy talkin