You example of "1 minute combat rounds" has NOTHING to do with association or disassociation; it's a measure of abstraction. The choice to "engage in combat" is directly associated because both the player AND character make the same choice--to engage in combat.
But under that test - equivalence of PC and player choice - 4e is associated too. Consider CaGI: the PC decides to impose his/her will on the situation; the player decides to impose his/her will on the situation.
My issue with "dissociated mechanics" is this: clearly there is such a thing as
metagame mechanics - that is, mechanics which contribute to or set parameters around resolution in ways that do not correspond to anything in the gameworld. But there is no actual evidence that using such mechanic must "dissociate" the player from the PC - as I've just indicated, it's perfectly possible to frame the choice to use CaGI in in-character terms.
Now you could say that CaGI has some extra feature - like the player deciding
not to try it again not for any ingame reason but because, as an encounter power, it has already been used and hence is not available again until a short rest is taken. But nearly all decisions that players make are constrained in various way by considerations that have no correspondence to ingame states of affairs - from worrying about being low on hit points (which for those who play with Gygaxian hit points is at least in part a metagame state), to wondering how much loot you should carry out given that 1 gp = 1 XP and your a couple of thousand XP short of a level, to worrying about irritating your fellow players, to thinking "I won't start a combat now because the session has to break up in 10 minutes time and we won't get it finished".
So if the criterion of "dissociation" is "player choice not equivalent to PC choice" then CaGI need not be dissociated. And if the criterion of "dissociation" is "player choice is influenced by considerations that are external to the ingame situation" then nearly everything in an RPG is potentially dissociated.
In addition to the ingame/metagame contrast, the "dissociated" discussion also gets tangled up with methods of resolution (karma vs fortune).
By
fortune resolution, I mean rolling dice to model probabilies. This is how 3E does skill checks sometimes - but not all the time. Sometimes you can take 10 - that is
karma instead of fortune - either you can do it or you can't.
Another place in 3E where there is an interesting approach of "sometimes fortune, sometimes karma" is in spellcasting. It is clearly true, in the fiction, that spell casting can fail if the caster loses concentration: hence the rules for inclement weather, riding in bumpy vehicles, taking damage, etc. But obviously there are many other ways in which a person can lose concentration - eg if casting while walking, you might slip on a stone, or otherwise foolishly twist your ankle, and lose concentration that way. But the rules don't model that - except in certain circumstances defined in the rules for combat, casting and Concentration skill spell-casting is resolved via karma rather than fortune: the player declares it, and it either happens (if the PC has the spell known and preparedn as appropriate to his/her class) or it doesn't (if the PC doesn't have the spell known and/or prepared).
It's interesting to note that this karmic spell casting resolution lets the player dictate features of the gameworld, including the actions of (very minor) monsters - for instance, when a player says "I cast a Light spell" and marks it of his/her list, the player has
also made it true, in the gameworld, that no mosquitoes or flies or bees or similar insects that might be hanging around are being so irritating as to require a Concentration check; that there is no pothole or puddle in the road into which the PC might slip; etc.
Come and Get It - especially in its pre-errata version - is karmic resolution. The player declares, in effect "I impose my will on my enemies" and the enemies comply. No dice roll is required. This is often characterised as "dissociated", as the player stepping outside his/her PC and making decisions for the NPCs/monsters, etc. But in my view it is no different from spell casting in 3E when no Concentration check is required. It is simply a situation in which the game rules allow the player's declared action to automatically succeed, even though it has an in-principle, ingame possibility of failing; and as a consequence of this karmic resolution, the player's declaration of an action can have implications for other features of the gameworld (eg the behaviour of NPCs; the behaviour of gnats and wasps; etc).
Those are the reasons why I don't think CaGI it is dramatically different from other, commonplace elements of D&D. The difference, for me, is not in the use of metagame mechanics, or the use of karma rather than fortune for resolution; it is in the application of those techniques in a sphere in which they have not normally been applied: namely, the resolution of martial endeavour.
), go re-read The Alexandrian's follow-up edit / revision of what "mechanical dissociation" means. By the revised definition, 4e is far, far, FAR AND AWAY the most "dissociated" edition of D&D (whether you like, dislike, or are ambivalent to that element of dissociation is purely a matter of personal preference). 
, you'll at least be starting from the same ground-level.
