I was arguing at work yesterday that relays have gain. Just because it's non-linear and electromechanical doesn't mean it has no gain. I mean a light switch has gain if you look at the power you can control with a finger. I think that silicon guys think gain must be linear or continuous, and be electrical in,electrical out. I think "gain" is much broader.
Am I right? Who owes who a beer? And on that note, could a carbon particle microphone be so constructed that instead of a sound wave input, I put a small headphone-style coil on one side and then arrange it to have gain? Does a coherer have "gain"???
Ah, the basics.