I see no reason a rabbit whisker would not work. I would place the whisker downgrade as far as practical. If that does not slow the car enough then go one more whisker a little ways upgrade.
I moved the magnet up again another 1/2 to 3/4 inch just past the crest and it improved uncoupling. Up from around 25% uncoupling to around 75% of the cars. With the MT magnet 50x7x3mm (measured) cars now separate at the very end of the magnet meaning to move it up further will require a longer or just more short magnets.
The idea is to open the coupler while there's slack, meaning while the car's still being pushed uphill. Then hold it open until gravity takes the car. If the peak of the hump is fairly flat, yes, that would require a long magnet.
I received my 1st set (of 2 sizes) of Neo magnets I ordered from China (a week earlier than the earliest estimate no less). I tried these Neo magnets and they pull the couplers together no matter which way I placed the magnets. Instead of uncoupling they do the opposite. I tried taking some pics but it doesn't show up as well as I would of liked. Compare those w/ the 3rd attachment. MT's magnet works, these super magnets do not. I even stacked two and other then it was stronger (probably because of height) the 'trip pin' still gets pulled inward, not outward on both cars. (Those magnets were not secured down. When the trip pin came close enough the magnet lifted up to the pin.)
Sumner...excellent video on how to use 'rare earth' magnets for automatic uncoupling. Note that the 'small' magnets are placed on the outside of the rails.
But that is for 'flat' switching which doesn't apply here. 'Humps' don't work that way. BTW, the 'poles' for these magnets are the top and bottom, not the sides or ends.
You still have to spread the couplers, not push their knuckles closed. In that way the process is exactly the same. And that's the only thing we use magnets for.
When the cars do separate they do it just like the prototype, gravity takes over afterwards. You are familiar with just how a 'hump' works? As long as both couplers move outward (in model RR operation), all is well. Roughly 75% of the cars do that, it's the 25% that don't for some unknown reason and it doesn't seem to be a Atlas vs MT issue. As a test, I temp placed two stacked magnets against the rail to see what happens. Nothing, the coupler never moved either way and that was with 2 magnets!
That's N scale?? In the upper right, is that a magnet? It seemed everyone else suggests outside the rails. I'm going to order another pair of MT magnets.
I received the MT magnet set (2) and discovered that the N & S poles are centered around the sides of the magnets which makes sense based on the direction the 'trip pin' is facing. On these small silver bar magnets the poles are on the top & bottom making them semi useless unless one positions them standing up om their side which isn't practical in my situation. The cost difference in nil. The solution was to cut one in half and extend the length which seems to solve most of the problems. It needed a longer reach past the crest which was the last problem. I still have a few exceptions which I'm dealing with the specific cars.