If you wait until you are at the end of an escalator before starting to walk, you risk stumbling badly, so we all learn to start walking before we need to - it's easier and faster for the body to adjust your gait to recover from the stumble, than it is to initiate a discrete movement to recover (whilst of course starting to walk).
On what I assume is the same basis, somebody advised me to alternate gentle light left/right pressure on the pedals prior to landing my tailwheel aircraft - if it starts to diverge to the left then I'm more likely to make a good recovery if I make the next rightwards stroke slightly longer and heavier than if I wait for the aircraft to diverge, then have to consciously recognise that I'm turning to the left and that I have to wake up my lazy right foot and mentally calculate how much I have to push it down.
I suspect the neuroscientists would explain this in terms of 'central pattern generators' which are fairly autonomous mechanisms in the spinal cord that control walking and many other repetitive movements. Whether similar mechanisms are enlisted for the manipulation of joysticks, I couldn't say, but I wonder whether making continuous movements is a fundamental part of how we fly - just that some people perhaps take things a little too far.