I know I'm resurrecting an old post here, so my apologies...finally had the time to read the entire thread.
As most of you know, I am a crossover trainer as well. Erin is the one who got me into clicker training. A lot of my feelings have already been expressed, so I won't rehash that.
One thing I did feel was missing here, though, was the emphasis on negative punishment. Not trying to put words in anyone's mouths, but the way I was reading the threads, it seemed that some thought that if you don't use a correction, you are ignoring the bad behavior or letting the dog blow you off. I don't feel this is the case at all. If I'm training a recall, for instance, and my dog blows me off to run and play in the yard, that is the end of the training session...he doesn't get to play, we go back in, no more training (which, to him, is a punishment as well as the not being able to play). Next time, we go out with a long line so he can't blow me off. I simply reel him in if he won't comply, or lower my criteria until I get the behavior that I do want. No physical correction, but he isn't permitted to not listen. That said, if he is blowing me off, it means, to me, that I haven't proofed that behavior well enough (at that level of distraction) and need to do more training.
I choose not to use physical corrections, as I don't see the need for them. I've been able to train everything i need my dogs to do without them. Using corrections when a dog is on-lead (or in an e-collar) is not a guarantee of compliance off lead or without equipment ,either...any dog
could choose to blow off a command, regardless of method, and i haven't found one method more or less effective in preventing that in general. Murphy is the dog that started in a prong, but is VERY easily stressed. It was too much for him. I can tell with his body language whether he is stressed or not...even without corrections. I do find for him, he is more reliable with positive methods, if for no other reason that he would shut down with any sort of correction.
For a dog that is trained using R+ and P-, the removal of the reward is VERY powerful. It does send a clear message since they know what it means. I used to think that it would take forever to train a dog using a clicker because I wasn't telling him what he is doing wrong...but I've found just the opposite to be true. I get the behaviors I want much more reliably and much more eagerly from the dog that I ever did using corrections. Do I know how to give them? Yes. Am I all positive all the time? No. I do use verbal corrections at times, and i have knocked a dog back who jumped up and hurt me, so I'm not perfect
I'm not slamming anyone's methods, as this is just a discussion...just wanted to emphasize that there are consequences other than corrections and that clicker training does not mean ignoring bad behavior (especially if the dog KNOWS the command). I only ignore the behaviors I don't want when I'm in the beginning stages of a new behavior...before it is even on cue.