The Dahl locking system looks promising. They don't sell to the USA yet, so I haven't compared ground clearances yet.
I saw a video of EZlock's Magic 360 crash test. The copy I saw was very low resolution so it was difficult for me to see what the point of failure was. It looked like the pin stayed intact. I think the top of the locking base may have deformed allowing the pin to slide out of the slot, above the internal locking mechanism. But as I said, I didn't see a good copy of the video so I could be completely wrong.
Using a plate, instead of a single bolt, seems like a good idea. It would distribute the loads on both the wheelchair bracket and the locking base. It would also act as the stabilizer to prevent the chair rocking fore/aft, side to side, and rotating.
The only downside that I could see would be a zipper effect, where the multiple smaller bolts don't load evenly. For example in a crash, all bolts should get the same shear force, but the torque caused by the chair would initially put the rear bolts under more tension than the front bolts. The back pair of bolts would fail first, then the next pair, then the next, etc.
https://dahlengineering.dk/en/products/dahl-docking-systems/