(I'm a game designer, so I can't help but respond to this as a designer first, and not primarily a player)

I wonder how much of the issue here is the rise of the abstraction of "gameplay loop" itself as a lens that shapes what gets made.

One of the things that can keep a game fresh is players being unclear on where the border of play is, or what the range of the possible is. When I was playing Mario 64, say, I really wasn't clear on what was possible in the game, and so one of the main pleasures of playing the game was encountering new kinds of interactions and new kinds of activities embedded in specific space that I didn't know would be in the game. Same experience with Legend of Zelda: Ocarina of Time. As a matter of fact, this was true for me when I was first playing through the original Half-Life and the original Metal Gear Solid as well, or Castlevania: Symphony of the Night. The boundaries of the possible were not clear, and I had to play to tease them out. There was something like an implicit promise (because those games were violating my expectations early) that I might see and do stuff I hadn't seen and done yet if I stuck with the game.

Refined gameplay loops with variation are certainly cool and a great and an important tool, and many games I love do rely heavily on them (like, say, Slay the Spire, or the original Doom deathmatch as a kind of competitive play, or Street Fighter 3). But my general sense is that the more designers think in terms of game play loops, the earlier the edges of a game design and limits on the realm of the possible become clear for a player in a subconscious way. In a way, this is similar to a player noticing early that they seem to have heard all the music a game is going to provide, or seem to have seen all the enemies or weapons early - they recognize they've found all the novel stuff they're going to find, and everything going forward is going to be re-combinations and permutations. But I think it's a little harder to reason about when it comes to a player discerning the limitations of what kinds of play they will ultimately encounter, because it's a bit more subtle of an issue.

There's something here about the aesthetics of open-ended discovery versus the pleasures of achievement, I think, perhaps in something like a fractal sense.

I think there's a lurking development tension here, too. Constrained variations within game play loops can often help constrain arbitrary interactions in game play code, and arbitrary interactions in game play code make systems harder to reason about, and balance, and ensure stability, and modularly farm out different tasks to different developers. So I suspect there are development reasons for preferring these kinds of designs as well.

Great comment. To add to this:

> There's something here about the aesthetics of open-ended discovery versus the pleasures of achievement, I think, perhaps in something like a fractal sense.

This is why INSCRYPTION was such a hit. You learn the game rules, and as you figure out the game loop, bam! you get a whole other game layered on top.

A game loop is necessary to create any game that’s not purely narratively based, but the secret sauce is to hide its seams well enough and integrate it with the game world (DARK SOULS and the cycle of death being a whole part of the story, Outer Wilds)

Or go meta and break the player’s expectations like INSCRYPTION did.

What you're describing I would categorize as "exploratory". Those games are fun and make good "big" games. The Portal series... early JRPGs... even the original Super Mario Brothers gave me that sense of not knowing where the boundaries are. You can have an exploratory game without much of a loop at all. But I also think exploration is unnecessary in some contexts. That's what I like about the field of games in general: they can look like a lot of different things and be good in a lot of different ways.