Oh and by the way? Dissociated mechanics exist along the N <-> S axis of Narrativist/Simulationist/Gamist. Gamism is mute on the subject.
This is fundamentally inaccurate on multiple points.
First, dissociated mechanics can be used to accomplish all kinds of things. Narrative control mechanics are
one thing they can do, but it's not the only thing. For example, the mechanics of
Monopoly are dissociated, but you're not going to find many storygamers getting all hot 'n bothered about the narrative glories of
Monopoly.
Second, dissociated mechanics can be used to achieve simulationist agendas, too. This is particularly true with the late-stage GNS interpretation of the term (emulating a given set of source material).
So really, once you boil it down, you realize that all three points on the GNS model are mute on this subject.
Of more use for this line of inquiry would be the Big Theory concept of "stances".
I suspect there's a certain mass of dissociated rules -- like hit points -- that should and would give immersion-sensitive players pause more often, except that we learn these rules very early in the learning curve.
The topic of hit points in these discussions is usually more misleading than enlightening: Prior to 4E, there was an interpretation of hit points (supported by most editions of D&D) which was associated but heavily abstracted. (And that abstraction was possessed of variable flaws depending on which edition we're talking about.)
Many people, however, didn't play with that interpretation of hit points. Instead, they played with an understanding that could be roughly described as "only the last few hit points count". This interpretation
is dissociated. It also served as the foundation for 4E's approach to hit points, which rendered it even more dissociated.
So someone who used the first interpretation says "hit points aren't dissociated", and then somebody who endorses the second interpretation says "no they weren't", and pointless bickering and confusion results.