I've read the discussion from
@Ovinomancer and you about 5e vs SP with interest. It doesn't really chime with my experience. I have played and DM'd all versions of D&D from Basic onward. I find 5e as concrete as any other version.
For example, resources - spell slots, concentration, ki, superiority dice, Hit Dice, inspiration, channeling, encumbrance, equipment lists, various class-specific pools, exhaustion, and a few recovery mechanisms. I have not found resource management to be "out the window" in 5e so there is a dissonance for me when reading that.
Similarly, the rules for 5e are tightly knit and concrete. It is a complete game system: there is little that cannot be managed with the help of one or other mechanic. If time permits, perhaps you could point to some B/X rules that can cast this "looseness" of 5e rules into the light?
This is really an aside from what the OP is meant to present, but I'll answer this quickly as it relates to Moldvay Basic and Cook/Marsh Expert (which will touch upon some other recent posts, including your revision of the Strahd AP...I think it is?).
There are a few key features that make Moldvay Basic the pinnacle of D&D Skilled Play:
1) There is no tome separation of player and GM. There isn't this massive mystery of GMing which undergirds play. Overwhelmingly, the game is table-facing (some have tried to use the epithet against it, as they did with 4e, that its basically a "board game").
2) Overwhelmingly, the game is codified in that text. The edge cases where a GM is going to have to go outside of the encoded play loop and action resolution rules to resolve something is remote. And where it does happen, its trivially inferable how to handle it (because the game's engine is light enough and consistent).
3) There is no "story imperative"/alternative playstyles (and all of the permissible GM moves/latitude around that) in Moldvay Basic that impacts (or can impact) the through line of play. The game is about one thing and focuses on that one thing; map/key/stock a confined obstacle course (the dungeon), create characters to defeat it, and referee the game neutrally to see if the players can and "what their score is" (the treasure they pull out of the dungeon) if they do.
4) There is a long and detailed history of discussion about the implications of leaving the dungeon on Skilled Play. When your game's skill paradigm is basically predicated upon testing guile, risk management, and resource management, the constraints of high resolution mapped and keyed obstacle course vs an obstacle course that is neither (refer upthread to
@hawkeyefan 's and
@loverdrive 's conversation regarding Strahd's lack of itemization of his inventory/resources he can marshal...this is one instance of the problem), at least sufficiently so to govern play at the level that Moldvay Basic does it.
Put another way, when you leave the dungeon, the horizontal, vertical, temporal, opposition resource components of play all suddenly become either profoundly more difficult to adjudicate/resolve and track or effectively impossible to do so (at least in the sort of constrained way that testing Skilled Play calls for). This is in part because map & key play becomes impossible outside of the dungeon but also, the sort of mechanics that Moldvay Basic uses to resolve that "dungeon as obstacle course" paradigm don't transfer well.
Expert suffers from this in a way that Moldvay Basic does not.
5) Ease of use + integration + beautifully conceived play loop. Exploration Turns + Exploration Action Resolution + Wandering Monsters + Required Rest + Monster Reaction + Morale + Encumbrance + Equipment Loadout + Gold for xp + Low Level (therefor bounded) play.
Outside of the Mapped/Keyed/Stocked paradigm above (which constrains permissible player moves and GM moves alike), these things above tightly govern the play loop of Moldvay Basic. The upshot of their integration is that they create pressure points on Skilled Play and boundaries/constraints for both players and GM.
Time and space tightly constrain the play space. The Wandering Monster clock + Required Rest pressure points loom like a Sword of Damocles. Exploration Action Resolution is straight forward, relatively punishing, and tightly rationed (and basically works as a dice pool mechanic for group tasks). Monster Reaction + Morale systemitizes key aspects of NPC response (which work with the paradigm of neutrality in refereeing). Encumbrance is abstracted and easy to use (therefore people will use it) and it integrates seamlessly with Equipment Loadout and Movement and has serious implications on Gold for xp work. All of these easy-to-use and integrated aspects of play work in service of the play loop/pressure points/score-keeping apparatus of play.
The more difficult things are to use, the more players will be inclined to elide them or outright eliminate them play. The less integrated things are, the more opaque their immediate and downstream effects are in testing Skilled Play. AD&D suffers from both of these problems. Moldvay Basic suffers from neither.
6) The impacts on play of level expansion. When your game is only level 1-3, again, the conceptual play-space (permissible action declarations, resources that can be deployed et al) is constrained and the implications of other aspects of system/play become considerably less amplified (the inverse is the case as levels accrue).