I can't see all the posts in these exchanges due to blocks, but from what I can see in quotes or the like, it seems to me that players going straight to ability checks to resolve traps or social interactions is not so much an argument for expediency in play as it is an argument for the DM to not present boring or simplistic content the players would rather skip.
Something like a decade ago on these boards I posted that the real function of Perception and Diplomacy in 3E is as a player-side reframing device:
Perception: The GM says "You see a room with XYZ," player makes Perception check, GM responds "OK, you see a room with PQR."
Diplomacy: The GM says "You see an angry person," player makesd Diplomacy check, GM responds "OK, you see a friendly person."
That said, I think my RPGing involves more checks than yours. On the weekend I ran The Dying Earth, and the centrepiece of that system is back-and-forth checks to find out who persuades whom to do what silly thing. And in my Classic Traveller game I follow the game rules pretty closely, and they call for checks in all sorts of situations.
This may be a function of system, but it may also be that I prefer systems that lean more towards "say 'yes' or roll the dice" rather than the "Middle Way" approach you've described in this thread. (Whether fictional positioning within the situation
affects the dice roll is for me a system thing. In 4e, Prince Valiant, Classic Traveller and The Dying Earth, typically yes. In Cortex+ Heroic/MHRP, frequently no (because of the way dice pools are put together in that system).)
Where I would say my approach differs from the "reframing" approach is that I see these checks mostly as a way to evolve the fictional position within a framing, rather than to reframe. I'm happy to accept that this, and related notions like the stakes "evolving" or "crystallising" as opposed to just "changing", are matters of degree rather than sharp distinctions. I still think it's a real difference, though.
In D&D rulebooks, I think the 4e DMG's description of skill challenge adjudication is the best account of this sort of thing. But one thing it's a bit weak on is what I think is the most important feature of this way of GMing: that there be no predetermined expectation (or even menu of expectations) as to the resolution of the situation - so the dice results, rerolls when they occur, etc, can be followed where they lead.
That means that my approach doesn't lend itself well to eg "house full of traps" scenarios, or "get the info dump from the quest-giver" scenarios. Nor anything which involves working through a pre-established map/key/event-list/etc.
EDIT: I thought I'd give a very short example of what I mean about checks to evolve a situation and its stakes, with reference to something I recently posted in another thread:
The last thing I can think of as dissapointing in a game session would be a year or so ago in Traveller: the PCs were in a domed city on a world with a corrosive atmosphere, with their ATV and some vacc suits. And I was pretty determined that I was going to get them out of the dome in their vehicle and suits to see what happened. And I did! - but in the course of that I discovered that the on-world exploration rules in Traveller are probably the weakest sub-system in the rulebook, bogging down with no guarantee of a resolution. I pulled out of it in the next session and the actual firefight in vacc-suits was excellent, including snagged oxygen hoses and shattered face plates and desperate crawling into air-locks and all the stuff that I'd been hoping for.
So the reason that the on-world exploration stuff
didn't work very well is that the rules tell you how to check for breakdowns, and terrain difficulties, and the like, but don't have a system for determining whether your get where you want to go and what it costs. (They assume that there is a map, and the players are calling directions, and the vehicle movement rate is applied. I don't like that approach in general, and in a sci-fi world-hopping game I think it's close to dysfunctional, though maybe good if the publisher has a lot of world map supplements to sell!)
What was good about the firefight was that the system supported tight resolution via checks at every point. Eg
I crawl up to the pillbox, which is a potentially risky manoeuvre in a vacc suit so triggers the check rule for that, check fails, so now something bad has happened (I narrated a snagged oxygen hose) and then as per the rules that triggers another, harder, check to escape the situation, that fails too, so now the hose has ripped off and the character only has the air that was still in her suit, and we resolve the consequences of that.
Or, after the pill-box is taken, the player of the smallest lithest PC declares
I squeeze through the slit, and again that triggers a risky manoeuvre check, which fails and so his suit is wedged - he pulls himself through anyway and so leaves the bottom half of his suit behind, and so we apply the damage rules for corrosive atmosphere (which had to be extrapolated from the rules for vacuum but that wasn't too hard) and then he makes a decision, inside the pill box, about whether to try and mix-and-match suit (he has a top half, the dead NPC in the pill-box has a shattered faceplate but an intact bottom half) or go through the air-lock that I narrate as being in the floor of the pill-box - he opts for the latter, and (as best I recall) I say "yes" rather than calling for a check because the "atmosphere" drama seems done and now it's about getting the rest of the crew into the enemy base and how that's going to play out.
I would say that while checks are determining in-fiction success, at the table they're not just about, or even so much about, player
success but also/rather modulation of pacing/drama - is the situation getting hairier, or under control? (This breaks down a bit when what's at stake is PC death, which can happen in Classic Traveller. But in the example of play I'm currently describing, we never got there and so I didn't have to worry about it!)