Maybe what they're talking about could be compared to, e.g., a similar method to teaching a placement sit? The one where you put a mild pressure on the dog's hips and wait until it moves out from under the pressure, even if it doesn't first sit all the way down. I suppose if used delicately and with good timing you could do something similar with an e-collar - i.e., used at a level that is just mildly annoying, not painful or frightening, and releasing immediately as soon as the dog does something that gets it a little bit closer to the right answer? I suppose you could arguably call it shaping if what you're doing is rewarding (ceasing the aversive) successively closer to what you want. And I suppose you could argue that a very very mild stim may even be less stressful to a dog than knowing there's a possibility of a treat and desparately wanting it but having to wait. Although I'm not sure I see what the real advantage would be to this kind of 'shaping' with aversives, over shaping with rewards? Is it that the proofing or teaching that the command isn't optional is more built-in, or something? Or is it the distance thing? (even if you click from a distance the dog still has to come back for the treat).

I can see it as a nice intro to e-collars so they aren't frightening for the dog, and that you could teach lots of things just as well on it and humanely, but they don't really make much of a case convincing me why it's better than an equally skilled person using a different method. It also sounds like it would be much harder to fade the e-collar as cue than it is to fade a reward or add a negative for a known?

But in any case, I can say that having someone like this on the end of an e-collar doesn't scare me, as some uses of e-collars do. Although clearly there are perfectly nice humane people using e-collars well, it sort of seems like if you look at it the other way around, if you're a nut boy are you ever going to be attracted to this tool.