The pathing and dungeon map questions here are useful, they’re one of the aspects of “OSR theory” that has gotten rather contorted through simplification - in that a fundamentally worthwhile concept and consideration has gotten distilled down to “Jaquaysing with looping corridors!”*
This is a bit unfortunate, because if dungeon exploration is a spatial puzzle for the players, excessive looping is almost as choice destroying as linearity. If everything is connected to everything else there’s no risk or decision about finding a path through it.
Of course, yes, maps are important, but without some signposting and clues in the keys they are mostly an annoyance for the players. Likewise maps matter a lot less if random encounters, supply depletion or something else doesn’t make exploring/moving itself a part of play with risks and cost… finding the right path only matters if it avoids risk.
A good dungeon designer will not only be thinking about paths, connections, and secret doors but also about ways the players can understand the dungeon and make educated guesses about how to find their way through. This can be anything from environmental clues (a draft indicates a nearby exit), indications of monster types and presence (signed stone near the dragon lair), old clues left by past expeditions (graffiti or partial maps) and indications of the dungeon’s history and rooms’ former uses (a kitchen will likely have storerooms and maybe a dining room near it).
Anyway, my point is largely that yes, making interesting maps shapes exploration and play. Understanding how players will move through them is helpful to design, but it is not only a mapping issue. I wish the Jaquays’ work on this were better understood in these early discussions, and the purpose of her complex (for 1976) keying and descriptions in Caverns were recognized.
For me looking back at 1970’s map design one sees fairly complex design. The LBB’s suggest complex maps (though perhaps smaller then the results might indicate), and the maps of the era follow suit. What early adventures often suffer from isn’t bad maps, it’s a lack of naturalism or more directly a lack of comprehensibility. What Jaquays’ and Gygax, though using slightly different tools, design brings to mapping and dungeon design isn’t the complexity of loops and sub-levels - it’s description that allows players to guess about the arrangement of the dungeon and content of keyed areas.
I wrote a couple of long articles on how the LBBs might have expected dungeons to look, and how Gygax designed commando raid style dungeons that were dependent on providing players information about the space and its defenders they can use.
On the LBBs Dungeon Design
On Gygax’s Design
*Note: “Xandering” appears to be a product Alexander feeling peeved that Jaquays wanted her name spelled right and then making some weird convoluted changes to support his book launch. I don’t think there was ill intent and it doesn’t detract from the utility of these posts, but I will use the term Jaquaysing cause I think it’s what she would have wanted.