I'm having trouble seeing where you come from.
That is, I can read the words and parse the grammar, but I don't understand what sort of play experience you have in mind.
If you have questions, I have answers. I think we obviously have very different play styles, so unless you find it highly beneficial to ask, no need to worry about it.
Of course not. If the player starts being the GM instead, s/he doesn't keep the daily either. This doesn't show that it is the character who uses the daily, though. It just shows that it is the player of that character who uses the daily. The authority to use any given power attaches to the particular role a given participant occupies - GM, player of character X, player of character Y, etc.
Right. Which is why I say it's attached to the character, and thus a specific-character mechanic. The character cannot consciously activate it, I agree.
As you yourself state, the daily attaches to whoever is running the character. That is not the character.
Agreed.
It is a real, actually existing participant in the game. The player, for whom the daily power is a resource.
I'd say that by that definition, all mechanics that the player controls are then, including the character. That's such a detached view from how I think I'd like to look at things that I cannot relate to it very easily.
Furthermore, this understanding of the power produces a coherent conception of the fiction, whereas your alternative doesn't. Your alternative produces the bizarre result that a being is exercising meta- or narrative control over his/her own life - which is fine for the Order of the Stick or, sometimes, The Simpsons, but isn't how I play my RPGs. The fact that it produces coherence seems to me a strong reason in favour of my understanding.
I think you may be misunderstanding my view, as this isn't the case from what I've talked about. I've said that the player gets to decide on a meta level, not the character. We're in agreement on that.
The character is affected by the narrative control feature (the daily power) attached to his character that the player activates.
It's similar to a GM's power to roll for wandering monsters, or decide whether or not a certain room in the dungeon has caved in after a heavy earthquake. These are powers that the GM - a real person - enjoys in virtue of occupying a certain role as participant in the game, analogous to the role of being player of character ABC.
Yeah, I understand that. Which is why I said earlier that it's a meta power the player activates ("I agree that the player activates the power from a meta standpoint, and that the character never thinks to activate it (it's a meta device

").
I don't even understand what this means. What does it mean for a PC - who exists, as a character, only in the fiction - to use an ability to manipulate or author that fiction?
It doesn't mean that. It means that when the player activates the meta ability attached to the PC, the PC then goes on to use the power -sliding a creature one square, or the like. While the daily power is narrative in use, the character "uses" it not by activating the ability (that's what the player does), but by actually sliding the creature one square.
Well, I agree with this. But given that "activate" and "use" are synonyms in this context - as far as I can tell - I don't see how it can be the case both that the player activates it, but the PC uses it.
Hopefully you understand somewhat better what I mean here now.
And this is the crux - it is, as you say, the player who can activate the power once per (fictional) day. So the notion of "using" or "activating" the power has no meaning within the fiction. So within the fiction there is nothing to be learned, explored or observed other than that the rogue, at least on occasion, pulls off some pretty fancy moves.
That'd make it dissociated to people that it disengaged from their role.
I don't know whaqt you mean by "testing a power in a vacuum". Given that you yourself have said that there is no such thing as the rogue consciously using the power (and by that I assume you don't mean the rogue uses it subconsciously); and given that the only coherent account of usage consistent with this seems to me to be that it is the player uses the power; I don't know what "testing" would consist in, let alone "testing in a vacuum.
You're saying that the rogue can be controlled in a narrative manner by the player to slide a creature one square, and the rogue is unconscious of it. I agree with that.
You went on to say that the rogue could do this more often if he had "the Low Slash and/or Positioning Strike encounter powers" or "a leader who has a power that lets his/her allies slide their enemies when they hit them." These were brought up to show how often the rogue could be controlled each day (by the player using the daily power).
To that end, you've added outside factors to the mix; instead of looking at how often the rogue can slide a creature one square, we're looking at how often he can do it with help, or with new powers. Looking at his powers without help from outside forces would be the vacuum I mentioned in my last post. And, even if we add the encounter powers, he can still only do it so often in a single encounter, no matter how long the encounter might last.
Testing would consist of looking at patterns to the narrative produced by using the powers. Testing in a vacuum would consist of looking at patterns to the narrative produced by using the powers without outside aid (from a leader, for example). If the rogue can consistently pull off one type of move a set number of times per encounter (or per day), no matter how long the encounter is, and this can be repeated dozens, hundreds, thousands of times, that would be the testing I mentioned.
So, the rogue is unaware of this narrative control. The problem in my mind is that a pattern can certainly still manifest itself, even though the rogue should have no grasp on the narrative mechanic whatsoever. That would mean that the mechanic could potentially be observed in-game, but the reasoning could not be learned, explored, or observed in-game. That would make the mechanic dissociated to anyone that it caused to lose focus on their role (lose immersion).
While rigorous testing need not be applied, just knowing that a mechanic works in such a way
can be dissociating in an of itself to certain players. I accept that it doesn't happen to you, your group, others in this thread, others on this board, maybe even others at large.
I mean, how do the inhabitants of the fictional world even frame the question in terms of "Did the footwork result from use of Trick Strike, or from use of Positioning Strike"? Let alone answer it.
They wouldn't. Exactly right.
As always, play what you like
