First bold: Again, it's important to distinguish between *lacking the ability* to do stuff (aka "impact the fiction via action resolution") and simply not being *great* at doing stuff.
There is a fundamental qualitative, not quantitative, difference between the two. You can absolutely still do stuff even if you're a poorly optimized character with junky stats.
Second bold: this is an odd place to focus. How often is conflict over action resolution occurring between the DM and the Players? This is a vanishingly rare issue in my experience.
It seems kind of like you're saying that players should want to have really good stats so that they can do whatever they want and not have to worry about the DM restricting them.
If so, that approach is very weird to me. Maybe I have misunderstood something, though.
Here's another way in:
When I read Gygax's PHB (and DMG too), I see roleplaying presented this way: the GM will draw up a dungeon (and perhaps a wilderness too); the players will try and "beat" it, using their PCs as the vehicles for that; those PCs have various sorts of mechanical options and constraints that matter at the table, that affect the "moves" that a player can make in trying to beat the dungeon; and most of those mechanical options and constraints also have some sort of in-fiction correlation, so that the player - in making the mechanically permitted moves - is also in some sense
acting as his/her PC in the gameworld.
In this sort of play, being a
good roleplayer means knowing how to make good "moves", in part by recognising, and trading on, the interaction/synergy between mechanics-at-the-table and events-in-the-fiction. Here's an extract from the Alarums & Excursions report of the original run of Tomb of Horrors at Origins, which illustrates what I mean (September 1975, authored by Mark Sawnson):
[T]here were no wandering monsters (damn few monsters at all, in fact), plenty of traps (too many) and very few experienced players. It was run by Gary's son, who devoted no effort to keeping the characters in character. . . .
Our elves reported no secret doors or traps. Ten more feet and out #2 and #3 fighters fell into a trap . . . We dragged them out. Should our Patriarch raise them from the dead? . . . With a hazy idea of saving the spell for later I ordered them dropped back into thee pit for later recovery. Neither the Paladin nor the Patriarch protested. The Dungeonmaster did not tell them they should have (both were neos.) No one suggested that we take their useful equipment along with us (one had a bag of holding.) At this point I ordered a Locate Traps spell used- a bit late- and we avoided two more pits on the way down to the end of the corridor . . .
We found ourselves inside a 10' square, 30' high room, without doors and possessed of 3 levers. At this point I announced that we were all driving spikes into the walls and standing on them. Various conditions of levers were tried. All three down resulted in the floor opening for a stimulating view of a 100' drop. . . .
. . . an orange mist door. Our sixth level cleric walked into it. A female anti-cleric promptly emerged and threw a curse. "I'm attacking her with my +3 mace" announced our sole living cleric operator. Gently the situation was explained to him. The Dmaster decided the curse had been hurled at the Paladin, Who picked the misguided female up and hurled her into the orange doorway; from which emerged our cleric with a sore jaw.
What the experienced roleplayers know, in contrast to the "neos", are things like:
* how to make sensible moves within the fiction (eg using spikes to protect against floors that drop away);
* how to deploy player resources effectively (eg casting Find Traps spells);
* common tropes, like a "trick" in a room that reverses sex and alignment;
* that certain alignments/classes (eg paladins, clerics) bring with them restraints on permissible moves (this is the alignment system not as part of a system of characterisation, but rather as an additional constraint on valid moves);
* etc.
As the mechanics of the game get more intricate, and more things are subject to mechanical resolution, stats and stat bonuses become more significant in being able to succesfully make moves. This is reflected in Gygax's comments about the importance of decent stats to a viable character. But the essence of
good roleplaying remains the same: working with the interaction between mechanics and fiction in order to make good moves that will let you beat the dungeon. (I think the place where Gygax spells this out in the most detail is in the closing pages of his PHB, just before Appendix 1. That discussion ends with the words "If you believe that
ADVANCED DUNGEONS & DRAGONS is a game worth playing, you will certainly find it doubly so if you play well.")
Now, I don't run a Gygaxian game, and haven't tried to since the mid-80s. But I can recognise core elements of the sort of roleplaying that I enjoy in the model that he puts forward: in particular, the idea of using the PC to make moves in the fiction (ie action declaration), which is then resolved by an appropriate application of mechanical systems and GM judgment.
In the 2nd ed AD&D PHB, on the other hand, I don't see anything of that sort. The character is not presented as a vehicle whereby the player makes move within, and thereby impacts upon, the fiction. The character is presented almost as a self-contained thing, which the player "gives life to" by establishing a personality, mannerisms, quirks etc. But the idea that the player should actually be using the PC to drive play is, to my eye, quite absent.
And, for me, this is the significance of the comment about stats. Gygax recognised that, given the mechanical systems and typical situations that characterise a dungeon-crawling game, a player whose PC does not have some good stats will not be able to fully engage. The PC is too likely too die, and the player won't be able to really take hold of the ingame situation.
It's possible to devise a game system where even a character with poor stats will be able to take hold of the ingame situation (eg by allowing metagame resources, or by adjudicating failure in a certain sort of way, or by being guided by player flags and signals in framing the ingame situation in the first place, etc). But 2nd ed AD&D isn't such a game any more than Gygax's AD&D is. It's not that the game envisages even weak characters nevertheless being effective vehicles for their players to engage the ingame situation and make their imprint upon the fiction.
Rather - and I think this is borne out by very many of the 2nd ed AD&d modules that I'm familiar with - the players are expected to provide characterisation and colour while the ingame situation is driven primarily, even overwhelmingly, by the GM.
And, as per my earlier post in this thread, that's the sort of conception of "good roleplaying" that I think is rather prevalent but that, for my own part, I don't really enjoy.