I'm a pretty hard core simulationist, and I believe that the purpose of game rules is to create a model of the world. For most things that we are modelling there should be a certain verisimilitude to the real world. That is the mechanics should model our own life experiences so that the results are familiar to us. You implicitly understand this when you write:
So I think we are in pretty strong agreement about what the process of play and the rules (ei, the fortune test and the resolution mechanics) are trying together to achieve.
Now consider the case you offered up as your first example, that of employing a weapon. When you employ a weapon against a target, isn't it the case that you get very immediate feedback regarding whether you hit or did not hit the target? You can in fact see that your blow landed, or that the arrow hit the bull's eye or otherwise know that you did well. So it's not at all wrong and indeed feels correct that a PC should see the result of his dice roll and have some feel for how he's doing.
But my contention is that not every real world test works that way - which is in fact something you seem to concede.
That's one example and one possible explanation but I don't think that is the sole explanation. Quite often in life you get things wrong, and sometimes you have a great feeling that you did well on a task which has a low DC and yet it turns out that you did poorly. There is actually an interesting psychological phenomenon called the Dunning-Kruger effect where the less skilled you are at something and the less well you do at something, the more confident you are that you did well at it.
With a little reflection I don't think it will be hard to remember cases where you thought you were doing well, only to discover you didn't. In my case I think of things like math exams where I made careless mistakes, programs I wrote that had stupid thoughtless errors in them, and playing 'pin the tail on the donkey' and thinking I was doing well when in fact I was wildly off. I don't think you can explain those just as 'misjudging the DC'.
It's my contention that in cases where the character can receive no immediate feedback as to how well they are doing, that the player should also receive no immediate feedback as to how well they are doing. So for a 'move silently' check or an attack roll or a climb check, sure I have no problem with the player rolling their own dice because as in real life, that character should be getting some sort of immediate feedback.
But in real life you have no way of knowing how well you are searching something. There is plenty of psychological experiments showing that people wildly misjudge how well they remember a scene or took in the details of something. How could you possibly know you were missing some obvious detail? I mean, haven't you searched a refrigerator before and felt you'd done so thoroughly, only to discover upon second look that the thing you were looking for was at eye level on the top shelf right in the front? Someone else pointed it out to you and you were like, "Oh that was obvious." It wasn't that the DC was higher than you thought it was, you just rolled poorly on your search check and missed an obvious detail. That sort of thing happens all the time. I would think life itself would refute your assertion without me having to.
Do you now have some idea how ridiculous this example is? Do you think I narrate an axe swing like that or ever roll the dice for a PC's ax swing in secret away from the player? To be quite blunt, you think you did really well in offering up this as an argument, but you actually "rolled a 1" and completely fumbled it. Yes, obviously for an axe swing a PC gets immediate visual feedback on the axe swing so there is no harm in giving the player the same immediate feedback. But that example only serves to show just how little you've actually considered this question.
PS: "i have left it with a good ferling, a goid sense of how i did" - Did you have a good sense of how well you'd spelled your response as well?
EDIT TO ADD - I misread the move silent reference thinking it was a case being put forth of a counter point to the attack roll. I leave my misread post here for evidence of my failure.
"
It's my contention that in cases where the character can receive no immediate feedback as to how well they are doing, that the player should also receive no immediate feedback as to how well they are doing. So for a 'move silently' check or an attack roll or a climb check, sure I have no problem with the player rolling their own dice because as in real life, that character should be getting some sort of immediate feedback."
But you make my point.
You are assuming they have no immediate feedback on their move silently check.
Are they deaf?
How many times have we seen in movies or other sources the character sneaking thru the woods step on a twig and freeze dramatically aware they just may have screwed up? How many times did they creep up the stairs only to have a stair creek and them stop knowing they just screwed up? how many times did they slip up and something gets knocked off... etc etc etc
"How much noise did I make" is the epitome of the player knowing how good his "roll to move silently is - unless for some reason they cannot hear the noise they make.
Don't they also have an idea about the background level of noise? Whether they times their run with the thunder or passing train or missed the timing?
What they are missing is the other side of the equation - the DC the sensitivity of potential listeners.
It is almost a text book case of "i can tell how well i did but i do not know how good the other guy is." which is my "you know the d20" but you dont know the DC example.
"
Do you now have some idea how ridiculous this example is? Do you think I narrate an axe swing like that or ever roll the dice for a PC's ax swing in secret away from the player? To be quite blunt, you think you did really well in offering up this as an argument, but you actually "rolled a 1" and completely fumbled it. Yes, obviously for an axe swing a PC gets immediate visual feedback on the axe swing so there is no harm in giving the player the same immediate feedback. But that example only serves to show just how little you've actually considered this question."
Actually i wonder if you have considered it much at all given the no immediate feedback on how well your move silently works example.
The point of the example is that in one case the Gm assumes no info can be gained - even for moving silently - and in the other they assume it absolutely can. That is a disconnect in presumptions i cannot justify myself.
"Can i as Gm provide (and should I provide) meaning ful feedback in the form of narration to reflect the difference between a 2 roll and a 19 roll?"
For a combat swing the answer is yes and that seems to be something we agree.
Most likely we are both also fine with that for saves and other combat related rolls.
But when it comes to "move silently" one of us says "yes and yes" and the other says "no" to at least one - even for move silently.
In my mind, why not use the same narrative sense to provide the same narrative feedback to the die roll when there are just as many possibilities of "stuff happens" to cover the 2 on search vs the 19 on search as there are for the swings?
"The closet was clean and well lit and easy to maneuver things around in - you are sure your search was thorough. 19"
"The closest looked clean and well orgnaized but when you started moving things stuff started falling apart, the light flickers every time there was a shift, and the search was the best you could get done in that time but far from what you would call thorough. 2"
"As you were looking back, a crowd came out and the wind kicked up so it while you failed to spot anyone following you at that point, you are far from confident. roll-2"
""The wind died down and the street cleared and given the turn of the road and clerar lines of sight you are pretty dang sure nobody is following you. roll-19"
The only reason there is "no feedback" in the cases where you as Gm "choose to not narrate the roll" is that you the Gm decided to not narrate the roll.
Again, whether or not the DC is met - different story. Whether or not the check matters - invisible inaudible trailer - someone with IRV seeing thru walls etc - different stories.
And no, i never feel good about my typing - being half blind with nerve damaged fingers - i dont GAF.
EIT i left out an answer - no i did not think you ran your games refusing to narrate the miss or treat that d20 roll as a unknowable thing. it was the point to show there is a disconnect between rolls that GMs choose to say "this d20 has an in-game narration" vs "that d20 has no in-game narration". As you ppinted out its somewhat odd to even assume an attack d20 would not be given narrative info. To me, it is equally odd to assume moving silently or lock picking or searching and a host of others do not have a narrative d20.