Hmm. So, I don't know the specificsfor that particular elemen but...while I liked Angband (and Zangband was even more of a "throw in the kitchen sink" thing), my own take on the philosophy behind the DCSS "remove elements of the game if they don't involve making interesting decisions" is positive.

Like, one thing people don't like is just doing...I guess "drudge" work in games? Like, in Cataclysm, one can result in loads of loot from dead zombies. Sorting through all of that can be a pain. The game could require you to do that -- which does add gameplay time -- or it could add features to accelerate that via automation, like searching for all nearby items containing a given material.

I can think of a lot of game elements like that. In some old games like the gold box Dungeons and Dragons video games from maybe the late 1980s, part of the game was walking through a dungeon, and as you went, having some graph paper handy and mapping it out. There'd be teleporters and secret walls and whatnot, and without a map, it was hard to make it around and fully search an area.

Over time, most video games stopped doing that. Exploration was still a thing, sure, but almost all video games had an auto-mapping feature that recorded seen areas on a map in-game.

Or taking notes as to what missions you had to do in RPGs and persons of interest. A lot of games from that era involved you needing to keep a notepad handy and joy down notes about people you needed to talk to and places to go. Today, most RPGs have some form of automatic tracking of missions and of important bits of information. Sometimes that involved even just providing waypoints.

I think that that's because, on the whole, game designers realized that transcribing maps onto graph paper just wasn't all that much fun. Nor was writing down each piece of information that sounded useful in case it might be important later.

What made something "fun" or not?

Linley Henzell's heuristic with Crawl was that if the player had to do something in the game, then it should be because he had to make some kind of meaningful decision, some in-game tradeoff. I should have the option to take this skill only if not taking it might be a bad idea. Use an item now or save it in the hopes that it might be more-useful later. If the player had to take an action and that action was a no-brainer, the only reasonable thing to do in that situation, then it wasn't interesting and the game shouldn't require a player to do it. Either the game should do it automatically -- as with DCSS's auto-explore or "remember locations of all seen items and let me search for them and auto-walk back to them" -- or just remove it as a gameplay element. Otherwise, it's likely to become one of those not-fun drudge gameplay elements.

I don't think that that quite captures all of video-game design -- like, in a World War II combat flight simulator, the long, droning flight out to drop a bomb on a tank might be part of trying to replicate the experience that real World War II pilots experienced, and it is that experience that is what the game is aiming for, though it involves little by the way of interesting decisions. But I think that it was legitimately a powerful insight into what gameplay elements ultimately had the potential to become "drudge" work rather than remaining fresh. Especially for a roguelike, which relies heavily on being very replayable, I think that it was a valid insight.

I won't say that every decision made by the DCSS dev team is the "right" decision. I don't know about the religion change they made there. But I think that the basic philosophy there contains a really important epiphany about designing games. I've seen game designers in other fields pick up on that and use that heuristic as well -- that is, it's been influential in game design.

DCSS is known among roguelikes for seriously paring down the game, to try to remove extraneous content that adds complexity without adding those "interesting" decisions. It aims for not having drudge work, and might also keep the learning curve on the game lower. Other roguelikes acquire a huge amount of functionality which makes the game bigger, and don't try to chop content. I like Cataclysm, but I cannot deny that while its simulation of stamina may be realistic, it can be a little annoying to manually lug an heavy item down a city block while having to stop and manually take frequent rests.