By rails I meant that the planes seem to follow changes in AOA very quickly. In AH film this can be witnessed well when the film maker had a spiky stick. The stick spikes create sharp changes in AOA and the plane seems to bounce very sharply after them when in real life I assume nothing else except slight changes in AOA and drag would occur from them.
I think that it's a gameplay concession - if the movements would be modeled authentically it would make gunnery much harder. The nose would bounce more than it does now as a result of stick movement.
Test some extreme low speed, low alt flying offline. Fly at stall limit and make a hard pull on the stick.
You'll notice that the AOA hardly changes at all, what you quickly bounce up and get a quick snap stall. My assumption is that a real plane would behave differently in the given scenario. The AOA would change violently droping the tail, stalling the wing, dragging the speed violently down and sending the plane to the ground in an unrecoverable spin.
In AH currently it's quite easy to go to stall speed, pull very hard on stick at stall speed and regain control with either increasing to full power or pushing stick slight forward to gain the speed back.
I made a small film offline where I flew at stall speed yanking the stick hard back, almost 90 degree angle to the ground at extremely low altitude. See it from external view and you'll notice how sharply the plane follows stick input with 90mph speed, full power. And then tell me I'd be able to do that in a real plane.
http://users.kymp.net/p404508a/film220.ahf