I guess I just don't understand how the adventures can be constitutive of the system. A really good writer can do almost anything with almost any system. Heck, Zeitgeist was made for 4e, but ported to PF1e and (AIUI) now 5e--yet it's still fundamentally the same adventure. How can that square with the idea that the adventures themselves are constitutive of the system? If official conversions to many different systems are not only possible but effective, what does that say? It tells me that a quality adventure will, in general, be quality at least somewhat independently of what system it was written for.
<snip>
Sometimes, a very well-crafted system will have excellent adventures written for it. For all its faults, AD&D1e was in fact pretty well-crafted, it's just absolutely terrible at explaining WHY it does what it does--and it, obviously, has numerous beloved adventures written for it. Sometimes, a terribly broken, flawed, exploitable mess of a system, like PF1e, has excellent APs written for it. Likewise, a terrible system can have terrible adventures, and a great system can have terrible adventures. There is no pattern here; system quality neither predicts, nor is predicted by, adventure quality. The two are mostly separate concerns.
Perhaps it depends what one means by "system".
By "system", in the context of a RPG, I mean
a means by which in-game events are determined to occur, or to put it in other words, the way that we (the game participants) work out what happens next.
In classic D&D, a huge part of the way in which we work out what happens next is by way of map-and-key adjudication. The adventures - the example of play found in the original books; the sample dungeon level with some examples of key and an associated example of play found in Gygax's DMG; the Haunted Keep and the associated examples of play in Moldvay Basic; and then the published examples of dungeons like B2, White Plume Mountain, Ghost Tower of Inverness, etc - demonstrate how map-and-key is done.
In post-classic D&D, adventures give us system in a different fashion - they show us how the GM can ensure that a story happens independent of the outcomes of the mechanical processes (mostly though not exclusively combat) defined in the rulebooks. Just as one example, the 3E adventure Bastion of Broken Souls contains advice to the GM on how to keep the story going if the players, via their PCs, eliminate the principal NPC antagonist (this is found in a sidebar headed "The Second String"). That's system being set out in the adventure.
To elaborate a bit further: 3E is, at its core, very similar to post-classic AD&D in its approach to PC building and resolution mechanics; thus, 3E adventures can be written essentially the same as DL-ish/post-Hickman AD&D modules. All development of the action - new scenes, the conduct and responses of the antagonists, the way conflicts unfold and resolve - is set out by the module author as something for the GM to narrate, with the role of the players being (basically) to provide the input that will generate the immediate outcomes of particular conflicts (mostly combats, occasionally other physical challenges like climbing a wall or getting through a door or crossing a desert). These modules will all have a weird relationship to information gathering activities, which in AD&D might be reflected via a reaction roll and/or bespoke rumour table, while in 3E might be reflected via Gather Information DCs - if the module is done well, ultimately these will not matter, because the adventure will need to proceed even if the players roll poorly on these info-gathering actions (and see eg the Alexandrian's "three clue rule" and its close cousin "node based design"). The architectonics, though, will be independent of player action declarations and set out by the adventure author as something for the GM to reveal to the players as the adventure is worked through. That's system.
On the other hand, as I posted upthread, it is not possible to write an adventure for In A Wicked Age. This impossibility is not a mere curiosity vis-a-vis the system: it reveals some of its profound features. The GM has no unilateral authority over initial set-up, which furthermore is not done as prep but is done by all participants as the first stage of play; and the GM has no unilateral authority over how the characters will initially be disposed to one another, which is a result both of collectively-established initial fictional positions, and of how "best interests", which are authored as part of the first stage of play, and by players as well as the GM.
There are a small number of published scenarios for Burning Wheel, and its interesting to see how they circumvent somewhat similar obstacles by relying on pre-gen PCs who have the right Beliefs to engage with the pre-prepared situation. I also have an example of this
here. But this won't work for In A Wicked Age, because of the different way its set-up works compared to Burning Wheel.
An example of a published scenario ostensibly for a particular RPG, but that actually contradicts the system and hence is literally unplayable as written - and its one I've commented on more than once before, in other threads - is Mark Rein-Hagen's
A Prodigal Son - in Chains in the Prince Valiant Episode Book:
One interesting feature of the Episodes in the Episode Book - which I have commented on more than once before - is that some are not well-conceived for conflict-resolution play, because they depend upon the GM breaking the connection between success/failure and win/lose in just the way Vincent Baker describes. The most egregious in this respect is Mark Rein-Hagen's "A Prodigal Son - in Chains" (pp 60-62), which contains such directions as "They [the PCs] need to capture and question Quink the hunchback and find out who he worked for" and "At this point you want Bryce to win over the Adventurers with his nobility of spirit despite his physical shortcomings" and "The Adventurers must now scour the forest to find Quink" and "Whatever happened, you need to have things end up with Bryce’s father, the duke, dead."
This is similar to a CoC scenario, or many TSR and WotC modules for D&D (including but by no means limited to Dead Gods). It is not compatible with conflict resolution.
Looking at why this is unplayable - whereas other scenarios that are quite elaborate in the fiction provided, like The Crimson Bull or The Blue Cloak or A Family in Despair, are eminently playable - reveals quite a bit about Prince Valiant's system (including but not limited to its mechanics).