I know all about having limited time to invest in changing arm action.
As you mention in the article, it's more realistic to assess a few critical areas in which the current action can be made more efficient.
Unless I have blocks of 1 on 1 time laid out over an extended period, I really can't see starting something I will probably not be able to finish.
I point out faults or inefficiencies that are closest to the first movement as I can find. As those things are corrected, and if the opportunities present themselves, we will work on the next thing, and so on.
I can buy into your assessment stage where the gun is not the first focus. Can the kid pitch, or is he just an arm? is a great question to ask.
With my son, he was inconsistent as hell for a while. Everything I tried to change would work at the start, but he would always revert to the original arm action / delivery, which wasn't horrible, but was not as efficient as it could have been. He was not getting his lower half moving fast enough and was trying to add velocity at the end with his arm. My fix was to tell him that if he wants to move his arm that fast at the end, then he needs to match that speed with his hand break and maintain it through release.
He was much more connected after that. He could develop a feel for his entire delivery and could tell if he was not right. He could begin to be his own mechanics coach. His velocity improved to the point where it was the rare kid on his team capable of catching for him.
I enjoyed reading your post, Lantz--as usual. Keep them coming.