Would chopped really slow conventional that much G? I'm not even sure I ever tried conventionals at my brothers.(I'm pure sine & conventional)
It still shouldn't drop the e-unit into a stall on the King's loco. (hey, there is a KFC boxcar for sale near me. Let me know; I'll pass you the info 
)
I read it all again as I had mixed up issues between locos. No biggie on what I wrote though.
The rev. switch being in the middle to work "better", points at some kind of issue there. I like to work by process of elimination... Make some 16g+ jumpers with alligator clips for that toolbox too
A jumper to ensure contact on the switch so the e-unit is basically 'hard wired' to "on" temporarily; does the issue clear up?
If so, it's the switch itself. Clean the lever contact, and the contact pad/rivet, check other connections.
Is the pressure of push nut/spring washer ok?; possibly re-seat the rivet to freshen the connection; wire brush the rivet top into the brim of the hole too, maybe fill rivet with solder for making a better pad; maybe pull the push nut and clean and shine it all.
The other e-unit thought is if the amp draw increases along that stall area, it could be curling any already weak e-unit fingers with heat build up. Again, jumping wires across it temporarily hard wiring it, if the issue clears, it's the fingers. (or maybe the drum or wires are broken/ shredded/ barely connected, cold solder, etc..)
Now, amp draw/track...
16g feeding a FEW 18g...much better. If too far apart with those , or real long drops, still might be limiting amps to the pulmore. (Should be good for can motors)
How large is the layout? (I think I know, but remind us please, 4x8ish ? "Overhead"?; around the room? 16g seems light for that distance imo. ).
And which DT&I do you have? One motor or two? I have the 8111 switcher (1 pulmore motor) and that has a 2-position e-unit, not a manual reverse switch.
Is the King a pulmore or can motor? (gives us a general amp use for comparisons)
If run alone or with all tracks going, is there an increase in issues? (track connections or wire/transformer may be too small.)
The coupler bit... If not under a warranty I'd confirm they work by jumping in power to the coupler coils. If they work, it's the board or remote. I'm not sure anyone has much idea exactly what's going on inside those yet. Try the loco couplers with the original power supply to rule out the chopped wave as the cause, (or an old pure sine transformer if one wasn't included with it.. even a car battery 
)
A spill or splash of "sticky" into the buttons is the most widespread remote issue overall I think.
Ive only had to clean/replace worn out buttons or replace LED&LCD screens in any remote of any kind ever, trains or not. (Increadible really. I never thought about it before today. Never even had a radio remote fully die on a toy that I recall)( I guess there are some knob issues with tmcc & thumb wheel on DCS...still "buttons" in a sense.)
