Personally, I think that "within 5 to 7 years" of a hobby that is 40 years old counts as early in the development of the game. Particularly when we're talking about official rule books that sold (and were expected to sell) hundreds of thousands of copies. They are a marker of the responses that had evolved in response to actual play experiences.
Early, indeed, and only a couple citations are being put forth as examples of (proto-)tailoring in what you're isolating (and seemingly trying to dismiss) as that early period. It also seems that you are overstating the importance of exceptions to rules. You need to understand that "fudging" is an exception to the rules which are by definition the standard.
But these are meant to be challenges for the players. Skilled players will, via detection (eg by dwarves or gnomes), via mapping, etc, realise that they are lost and (as per the advice in Gygax's PHB, p 109) will make finding their way out of the dungeon their number one priority.
This shifts toward the OP discussion, just to be clear and avoid conflating the two discussions, but even on page five of the (O)D&D's third booklet in the sample underworld the third entry states, "This area simply illustrates the use of slanting passages to help prevent players from accurately mapping a level." There's no guarantee that players and their PCs will know they are out of their depth.
But, here's a bit of proto-tailoring on the same page in a later example, "The combinations here are really vicious, and unless you're out to get your players it is not suggested for actual use. Passage south "D" is a slanting corridor which will take them at least one level deeper, and if the slope is gentle even dwarves won't recognize it. Room "E" is a transporter, two ways, to just about anywhere the referee likes, including the center of the earth or the moon. The passage south containing "F"is a one-way transporter, and the poor dupes will never realize it unless a very large party (over 50' in length) is entering it. (This is sure-fire fits for map makers among participants.)"
Note the potential for adversarial GM vs players (poor dupes?
) gameplay as being a distinct possibility in the designer's reckoning? That's rather metagamey and speaks to the OP discussion here. There's the transporter, btw, that I think someone else mentioned as not part of (O)D&D.
But, I think the point here is that the players can often have no idea they are even out of their depth, and far deeper in a dungeon than they know. They might come realize it when they lose a few of the party to creatures they had never previously encountered and then make the priority to get the heck out as soon as possible, only then discovering that at some point they took a one-way trip. Or they might never figure out how they came to be where they are, figure out they are beyond their ken and struggle to find a whole new egress.
Because of your own predilections, you take the idea of challenging the players as an ideal, as discreet and defined challenges for which there are specific solutions or, at least, thought to be solvable. But this is at the very heart of the difference between sandbox play and what came later. Challenging players can simply mean creating the circumstances whereby players, through their PCs, can find themselves in PC-life-threatening situations, and letting them figure their way out while the GM might have no idea if it is even possible.
To shift back over to the player authorship discussion, how effective is a passage slanting downward into unknown depths if a player can simply put crates at the end and climb back up and out through the trapdoor that appeared at the player's authorial direction? Can the player justify that since he just killed ten healthy human guards they must have had crates of supplies and also a trapdoor through which to bring them into their guardroom? Does this early period of D&D allow for such authorship? At what point in the development of D&D (leaving aside other RPGs and storytelling games for the time being) does such a scenario become the norm for D&D rather than unusual for gameplay?
And the evasion rules from the early game, in combination with the convention that placed monsters don't make forays out of their rooms, will make that feasible -
You need to understand that "wandering" monsters are one conceit that creatures do indeed make forays out of their rooms. Furthermore, "As a general rule there will be far more uninhabited space on a level than there will be space occupied by monsters, human or otherwise" indicates that the distance between lairs has some sway over something that has treasure to guard will, if it hears something dangerous wandering around or fighting in the distance, will come to investigate. Also, "At the end of every turn the referee will roll a six-sided die to see if a "wandering monster" has been encountered. A roll of 6 indicates a wandering monster has appeared."
That's an average of once every hour spent underground that something making a foray will be encounter. So, think of it this way. If I am a monster in a dungeon guarding some precious treasure, why would I go investigate every hourly sound and risk being killed by something more dangerous than myself? When I need to feed, am I not better off stalking something on my own terms than rushing toward the sound of a battle, battles generally being between two things with the means and will to fight? If I lie in wait and don't make noise then, as per the above standard conceit, something is going to wander by each hour and I can increase my chance to surprise it by not making noise myself. Granted, this precise line of reasoning isn't spelled out but it certainly follows from what is written.
(. . .) they can run from wanderers, dropping treasure or food to distract pursuers, and make it out.
That's certainly one player character strategy to avoid getting eaten.
Doing this sort of thing in 3E is much harder (. . .)
So, on the one hand we're dismissing the standard gameplay of the first five to seven years from which only one or two exceptions of that standard are cited as so early it is a blip in the timeline of RPGing but in the same post we're also reaching forward over 25 years later to cite rules where the early proto-tailoring is finally fleshed out. As I have pointed out, there are turning points and a long development period.
Nowhere do either Gygax's DMG or Moldvay talk about sand-boxing as a mode of play.
You only need to label something as such if it isn't the default by design, as pointed out in the examples from which we've collectively only cited a few exceptions from the first five to seven years. There's no need to call ice cream vanilla ice cream since, until there is a chocolate, vanilla ice cream is simply called ice cream. The labeling of sandbox play doesn't come until there are other forms of play beginning to become popular.
And there is even the remark that "players involved in the outdoors someplace will either have to come home to 'sit around' or continue adventuring in wildernesses and perhaps in some distant dungeon as well (if you are kind)". What is the kindness that Gygax has in mind? Presumably, the GM placing a dungeon for these players, whose PCs are ahead in the timeline, to explore. Obviously that is not Forge-ist scene-framing, but it's not pure sandboxing either.
The kindness to not simply play in the default style? You seemingly recognize that by your final phrasing and also understand that this argument you are making is pointing out an exception to the standard mode of play. You're not acknowledging this directly but your phrasing here shows you do inherently understand it to be true.
My reason for makig these points is the one that I stated upthread - that the approaches to authorship, backstory management, scene-framing, etc that are identified with a range of different RPGs (including Forge ones) didn't come from nowhere c 1995.
You're arguing something in this one quoted passage that no one else is arguing against but the leap you are making from that seems to simultaneously need to dismiss the first five to seven years of RPG development as insignificant (because citations that support your arguments from this period are thin or overstated) and act as if the rules thence forth fully support modern storytelling conceits rather than being introduced over time as exceptions to standard gameplay (which everyone who has discussed it in this thread seems to agree come from how some GMs and players, including the designers who still played the game, might have played at their own game tables).
They have their origins in real problems of RPGing,
Here's where you go even further off the tracks and show your belief that your idea of fun is everyone's idea of fun, as you are identifying what you see as problematic then casting it as universally problematic. I won't join deeply into this tangential argument as the premise is obviously faulty but I will point out the following. What you term as "real problems" are just how folks played (and many still do and some just starting with new games like DCC do) but doesn't seem to agree with your own way of playing. I think this underlying bias hinders your ability to objectively examine the development of RPGs, RPGs with storytelling elements, and full-fledged storytelling games.
