As I understand it, this is the issue that motivates [MENTION=386]LostSoul[/MENTION]'s 4e hack - he wants more "rightward arrows" (from the clouds (=fiction) to the boxes (=mechanical gamestate(?))) than he finds in 4e as published.
I think that skill challenge resolution, as written, requires rightward arrows - the GM has to frame the initial situation, and then reframe as part of each new skill roll (PHB p 259; DMG p 74):
Your DM sets the stage for a skill challenge by describing the obstacle you [the player] face and giving you some idea of the options you have in the encounter. Then you describe your actions and make checks . . .
You [the GM] describe the environment, listen to the players’ responses, let them make their skill checks, and narrate the results.
I interpret the plurals here as distributed, not collective - ie after each description a player responds, makes a check, and a result is narrated which provides the new environment to which a player then responds - because the other reading - describe the environment, let the players make X checks without any connection to the fiction, then narrate the overall outcome of the challenge, (i) seems to produce a crappy game and (ii) is at odds with the examples of play that are found in the DMG and RC.
Because of the role of the battlemat and tokens/minis, I think that the role of the fiction in 4e combat is more contested. Some people think that the map and tokens are a represenation of the cloud. But obviously they are also part of the mechanical gamestate, and so are boxes.
I think how 4e combat is experienced may depend a lot on whether, for any given group, the stuff that is drawn on the battlemap is first and foremost
fictional stuff - trees, rubble, fog, walls with doors and windows, etc - or first and foremost
mechanical stuff - cover, difficult terrain, obscuring terrain etc. Perhaps in part because my maps are fairly sketchy and my group uses board game tokens rather than miniatures or even WotC's picture tokens, I think that the fictional stuff prevails. And this is reinforced by the resolution of interactions with it that involve rightward arrows and not just manipulating the map - like climbing walls, overturning furniture, opening or closing doors and shutters, etc.
This in part relates to Vincent Baker's comment #4 on the blog you linked to:
There are a couple of places in the game where there are supposed to be rightward-pointing arrows, but they're functionally optional. I assert them, but then the game's architecture doesn't make them real. So it takes an act of unrewarded, unrequired discipline to use them. I suspect that the people who have the most fun with the Wicked Age have that discipline as a practice or a habit, having learned it from other games.
To an extent, my group has habits developed playing other games (mostly D&D and Rolemaster). But there are also aspects of the 4e architecture that generate rightward arrows - the rules on damaging objects, for example, make it clear that keywords (like fire, ice, teleportation etc) have fictional signficance. A tree can be set alight, for instance, but a stone pillar can't - so here we have a rightward pointing arrow, from fiction to mechanics, that is not just boxes (in the form of a cover symbol on a map) to boxes. Icy terrain can be used to cross a river, whereas a grasping vines spell that also creates difficult terrain probably can't. And so on.
[MENTION=6679265]Yesway Jose[/MENTION] is suggesting that this sort of approach to the game is not common. I don't know whether or not that is true, but I think approaching the game as a purely boxes-to-boxes exercise, or boxes-to-clouds plus a bit of clouds-to-clouds ("improv drama linking the tactical skirmishes") requiers ignoring things like the signficance of keywords + fiction to action resolution that are expressly called out in the game rules.
Anyway, I'm not sure how (if at all) this relates dissociation, but I do think it's an important issue in game design.