In a similar vein, the presence of alternative choices doesn't vanish once the PCs engage with one of the caverns.
<snip>
But if we lay all such considerations aside and hypothesize a beginning group that simply chooses at random and bulls their way through each complex room-by-room without much thought or interaction, the interconnected layout of the caves and the fact that they
could have chosen a different path (even if they would have done so at random) may still teach them an important lesson. (Which will take either the form of, "Holy
, we died. Let's do better scouting next time." or "We sure got lucky that we came up on that ogre from behind instead of via the pit trap. I wonder if we could make our own luck next time?")
I've cut some stuff in this quote not because I disagree with it but because I think it speaks for itself.
I just wanted to pick up on these parts of what you said because, for me at least, they raise the question of "What is the dimension of significance in respect of which player choice should be encouraged?" One possible dimension - the one that seems to me to be at work in what I've quoted - is that of
operational optimisation. And for the reasons you say, the Caves give the opportunity for the players to make operationally more or less optimal choices.
I imagine that the KotS allows players to learn, as they work through it, how to make
tactically more or less optimal choices. That would be another potential dimension of significant choice. I don't have an opinion on whether it does a better or worse job of this than B2.
Belly of the Beast, which I mentioned upthread, allows for meaningful choices in the domain of politics/personal morality - choices about which factions to align with, or oppose, for what sort of reasons. I will voice the opinion that Belly of the Beast does a better job than either KotS or B2 of permitting players to make meaningful choices in
this particular domain of signficance. I know KotS offers something here, because of the ghost of the dead knight, and (I think) some Orcus cultists in town. I don't think that B2 offers any more of this than KotS, and as written - without very extensive embellishment by the GM in respect of the Keep at least - may offer less.
Maybe Mearls just doesn't care especially for operational play. I think that that would fit with The Shaman's view that he doesn't get old-school play. I don't think it follows that he can't design adventures.
(I know of one Mearls adventure that does focus on operational play - it comes bundled with the Ghost Machine and I think is called Swords Against Deception. It focuses on infiltrating the island on which a cult is living, and then entering their fortress and killing their leader, without getting caught. Even this differs a bit from classic operational play, in that the goal is not open-ended looting operations but a single successful raid with a tightly defined goal.)