Tim did not say what is the purpose of the negative pin although to be fair I did not specifically ask. His answers were the typical, vague, non-technical responses I have come to expect from him. No warranty, risk of fire, blah, blah, blah. The one nugget of information he did supply is that there is no LM over-current protection in a blue wire / ground pin capacitance application. Since that was the key piece of information I was seeking I stopped there. I make it a habit not to post email conversations publicly online but you aren't missing much. That really was the only juicy answer he provided. He did say he is aware some people are connecting KAs in this manner but quickly followed with another bout of no warranty, at your own risk, etc.
When you think about it the answer could not have been any different. The ground pin is immaterial. The root problem is huge capacitance on the Vcc rail at the load. Absolutely the worst place to put it in any electrical circuit.
One good use I can envision for the ground pin is to add a small non-lethal amount of capacitance. Just enough to keep the LEDs lit and the sound system from resetting when momentary loss of rail contact occurs but not so much as to fry the output transistors from a short circuit. Much like what other DCC sound decoders do with their external capacitor. This wouldn't do anything for the motor function but, as I understand it, momentary sound resetting is a common annoyance for some. It may fix that problem.
Tim isn't going to reveal granular design details. Nor should he. He has a business to protect. We can only speculate and deduce from basic electronics facts. We suspected, and Tim confirmed, the blue wire / ground pin is not a safe place to connect a motor KA. All this exercise did was confirm you guys are doing it the right way with your homemade KA circuits. At least on LM1, LM2, and LM3 units. Who knows what is in store for LM4s.