4e .... where the rules in large part negated the need to have a social contract in place for optimization, conflict, and class balance "trouble."
If, as a designer, you view the D&D ruleset as primarily a vehicle to combat resolution, then removing the need for that social contract probably seems like a high-minded, absolutely necessary step for the game to evolve.
At it's core, 4e's seductive undertones are, "Don't be beholden to the whims and fancies of DMs and players. Build the character YOU want, and it will work. Never feel useless, never let those Wizards and CoDzillas rule you again."
It's a powerful, persuasive argument to the right kind of players and groups.
I don't particularly see what combat resolution has to do with it. 4e has more robust social resolution mechanics than any earlier edition of the game that I'm familiar with.
And the idea that players should be able to build their PCs, and play them by the rules, even push a bit, and the game not break - that should be a basic design goal for
any RPG in which it is assumed that the players, rather than the GM, have primary responsibility for hurling their PCs into the throes of action resolution!
Anything less, in my view, leads to insipid, GM-controlled illusionism (or blatant railroading), in which the players' main contributin is simply to add colour by emoting their PCs and describing details of their actions that have little actual bearing on action resolution.
(I'm not saying that a game is fundamentally flawed if, for one group, with their own table preferences, one sub-component has to be excluded to make the game work. I've had this experience with Rolemaster - eventually, my group discovered that we couldn't make RM work if we didn't just ban much of the divination magic. But that still leaves the core of the game intact and working.)
you really have to have one or the other--players who agree not to stretch the limits of the rule system, or a rules system that keeps the players within highly codified limits of "stretching."
What survey of functional RPG designs are you basing this claim on?
I'm not saying that my survey is even approaching comprehensiveness - but I don't think that Classic Traveller, or Runequest, especially requires players to either agree not to stretch the action resolution mechanics, or alternatively to be "highly codified" in their limits of stretching. (In Traveller, there is the whole "battle armour" issue, but that is going to be marginal except in a certain specific sort of campaign.)
Going into more abstract systems, I don't think HeroWars/Quest, or Maelstrom Storytelling, is going to break when pushed.
And there are all sorts of buffers you can build into a rules system to ameliorate the pressure that players bring to it - like giving them a reason not to always want to bring all their dice and bonuses to bear, and like setting stakes that are less than abject failure or death. (Call of Cthulhu is one example of a classic game that I think ticks both these boxes.)
It is a distinctive feature of D&D, I think, that it nearly always gives players an incentive to maximise their bonuses, in part because the stakes are always so high, and then has a tendency to break under that pressure. (Of other RPGs that I'm familiar with, Rolemaster probably comes closest to replicating this feature of D&D.) I'm currently GMing a 15th level 4e game, and it is highly noticeable that despite a lot of pressure from (at least a couple of) the players, there is only one ability (a feat from Dragon that lets the fighter immobilise marked targets whom he hits with a basic attack) that is currently on a house-rule watchlist.
There are games and mechanics that attempt to actually explain the mechanics in-game; D&D is that sort of game, and most of the mechanics are like that.
<snip>
There are games and mechanics that give players explicit narrative control over the game and their characters. When the first kind of game introduces narrative mechanics, they tend to be universal--everyone in 4e has action points and healing surges, for instance.
Yet 4e martial dailies look almost like an attempt to give one power source narrative mechanics while other power sources have internally-consistent or simulationist mechanics--and then they act like those narrative mechanics are simulationist.
I don't see what's objectionable about using narrative control metagame mechanics to balance ingame abilities. I understand that the Buffy game does this. And HeroWars/Quest is a game in which metagame mechanics can be spent either on character development or on boosting die rolls, which is somewhat analogous to choosing between process simulation and narrative control. And
I published an idea along these lines in a HARP/RM online fanzine in 2007, based expressly on the idea that a PC could either opt for metagame/luck based success, or ingame/skill based success.
and then they act like those narrative mechanics are simulationist.
Who are "they" in this clause? Nothing in the rule books calls out martial dailies as process simulation abilties. And treating them as essentially metagame abilities is a pretty obvious option for anyone familiar with the idea.
The rulebooks themselves somewhat gloss over the issue, in much the same way as D&D traditionally glosses over the issue of what hit points represent (if anything).