You made an interesting observation...I can't tell you how many times Ive told my students " If you can't change the dog, change the environment..." If the dog can't open the refrigerator door, put a knotted rope on it for him...
On your pup too wound up to sit...it is a common problem in teaching a passive indication ( sit) to a dope dog.
The way I do it is an extrapolation of basically all dog training.Again, its incremental. Stuff the toy ( i use a kong whenever possible) with dope, put it under your foot and let the dog see you do it. He's all excited and not listening to a word you are saying, scratching at the foot, jumping around etc. In a quiet calm voice say "sit" and repeat it...and nothing else at about 3 second intervals. Eventually the dog is so frustrated that he's willing to try anything, which includes listening to you. He sits....and needs to get immediate big time positive reinforcement. About three times and he is starting to get a clue that the quick way to his toy is to sit. In this phase he can see, smell, and touch the toy.
Next you go to a line of three milk crates. Put the loaded kong under one and start him on the line search. He will see/smell the toy and go nuts trying to get it. Again, put your foot on the crate and say 'sit' quietly. He catches on pretty quick! In this phase he can see and smell the toy.Putting your foot on it is a piece of the scenario he recognizes and it jogs his memory of what to do quicker.
Next is a line search with 3 or 4 cardboard boxes, loaded kong under one. He recognizes the line search scenario, can't see in the boxes so his instinct to smell for it kicks in. Again he might get crazy trying to get it, but again put your foot on the box and elicit a sit.
The bottom line is you are conditioning him to sit in order to get his reward. Neurologically when that conditioning is in place it is called a limbic system response.
Contrary to what you said, you CAN back up from any phase of that, and we often do. Dogs with years of experience will sometimes decide they'd rather scratch than sit, or do something else. At this point I literally go back to the beginning, kong under foot, and elicit a sit, with heavy rewards. Because the dog recognizes this scenario usually about 5 minutes of this 'tuning' dries up the issue. Enforcing the command, at least in the way I train, is counterproductive. If he doesn't have fun doing it, and do it of his own volition his performance can be erratic. Erratic performance does not fly when applying for an affidavit for a search warrant, or establishing PC on a case thats going to end up in court. Basically the dog has to be nuts for the work ( or in this case play..) or he isn't going to make it. We like the kind of dog that will do anything including following simple commands to get his reward.
