Well, it's a long story, but I read the posts before I decided that my computer was so unresponsive (jammed adware script, it was) that I sicced my anti-virus onto a complete computer scan... two hours later I replied to the posts I had read, without realizing that the real problem had been posted in the interim. Namely, that the control of a conventional engine was involved and the e-unit would not release due to the amount of energy remaining in the left handle circuit. So I unintentionally minimized your problem.
My club layout could not handle conventional engines, only DCS and TMCC, so I did not think of conventional. I had said all the power wires were bundled together. This was a simplification. Actually, all the hot wires were bundled together, and all the common return wires were bundled together separately; the two bundles were about six inches apart on the same route. It is this separation that caused our problem, IMHO... you see, each of our four circuits had 50 square feet of air between the out and back conductors of that circuits. Now an air-cored transformer is not very efficient of space, but that does not mean that it is not effective at times. Here that air-core is equivalent of about 7 square inches of iron core-- that's about 3 times the core area of a postwar ZW. Granted it's only a single turn, but 3 turns on the core area of a ZW is about 1 volt.
Well, that's at 60 cycles... with DCS you can say that you may be transferring power at 400 cycles. In that case, the iron core equivalent is about seven times more effective (400/60), so you may be transferring seven volts, which definitely will interact even with the postwar electromechanical e-units.
So, what does that mean? Well, if you are using star wiring and the OGR wire with an outer jacket over the out and back conductors, you don't have much air between out and back. But not all do that. Some say a lot of wire can be saved if the common return is made thru the outside rail and just one connection to the transformer(s) (or the load side of the TIU(s) at a point where the outside rail is closest. With can motors and cruise control of DCS or TMCC or Legend locomotives, you may not notice any problems. But this arrangement can result in the out and back path of the traction current being separated by a lot of air. Look into this by making a scale drawing of the current path (on squared paper) out and down the track and back along the outside rail. Whatever wiring layout you have, look at it in these terms and give it some thought.
--Frank
