Well I'm not comparing the model world to the real world in the sense that ALL the forces are comparable.
Of course one can model area a lot easier than one models mass. Our rolling stock would be LOTS heavier where it the same relative weight.
This fact doesn't change what forces are in effect. It only effects how and where they work..
I stand behind Newton until he's proven wrong
At the real speed our stuff runs at centrifugal forces are negligible. The only significant forces are wheel friction caused by the inner rail being shorter (making the inner wheel want to turn slower) and the drawbars trying to pull the cars off-track towards the center. This friction and how it aggravates the required pull is the real issue, and it is not helped by super-elevating a curve.
With a 20 inch radius curve 1 g is reached at 5 mph. That's a 435 MPH scale speed just to generate 1 g of that fictitious centrifugal force!!! These forces follow a square law, so at 2.5 MPH (217 MPH scale speed) we are at 1/4 g.
At 70 MPH scale speed g-forces are under .026g's. Every time speed is halved, g-force is quartered.
The other consideration is radius, and we can't get 1/2 inch out of the radius with any bank angle. The tracks are not wide enough compared to the radius for this to be meaningful. Clearly anyone proposing the idea that banking the curve increases effective radius any meaningful amount is absolutely wrong. Changing radius is also a linear function, not a square, on g-forces. Halve radius we double g-force, double speed we quadruple g-force.
Super-elevating clearly would make things worse at the speeds our stuff runs at. The dominant force by far should be draw-bar pull caused by friction as the wheels with common axles try to rotate at different speeds, not forces caused by speed.
Banking would tend to push the cars towards the center of the radius, and since the draw-bar is above wheel contact height the pulling force would add to the banking force and tend to roll the cars toward radius center even more!!
Now an opposite bank, where the cars tend to lean outward (outside rail lower in height), might help keep the cars on track better at slow speeds. The draw-bar would then pull the truck DOWN into the rail, not lift them out. Of course backing up would get worse because the draw-bar would be in compression.
As we increase draw-bar drag, we want to reverse the normal bank (or lower the draw-bar height above the wheel contact point) when pulling around a curve. As we run the train faster, we need a more typical bank to aid centripetal force.
Anything running around curves at 400 MPH scale speed? Or do you back around the curves? Then by all means, bank them.
Tom