...It seems bizarre to characterize wanting rules to lead to predictable results as "gaming the system".
Please reread what I said. I did not say that wanting predicatable results meant you inherently were trying to game the system. I said stealth and surprise are areas where precision CAN be abused and busted - and that the ones that complain the most tend to be the ones that want to game the system. I can say every orange is orange, but that doesn't mean everything orange is an orange.
On the contrary - reliable and transparent rules which don't require DM hemming and hawing typically result in players very much not "gaming the system" in my experience of many TTRPGs.
This is the oppostie of my fairly extensive experience. When you have reliable and transparent rules for complex situations, you end up with a rules lawyer playground. They do things like point to the clear rules that show they can 100% always sneak up so long as they combine these 8 elements. In systems that went that direction, I allowed it as a DM - but it stopped being a good game with thrills and story as players tended towards looking for opportunities to show off the rules lawyer trickery, regardless of how nonsensical it was.
D&D is a role playing game. Part of what works in this style of game is some uncertainty. It is why we have dice in the game - to add some chance and an inability to know for sure how things will play out. It adds to the thrill and takes the story out of the hands of the DM - to an extent. It provides the thrill that makes it an adventure rather than a scripted event.
Players just want an ambush to be an actual ambush - hardly unreasonable. 5Es Surprise system is almost uniquely bad, and it's bad for everyone as it provides results which are unsatisfying, counterintuitive and often downright confusing, whilst not even being particularly fast.
I'm confused now. The 5E system is almost entirely up to the DM to determine how to resolve. There is almost no structure to it. When the game doesn't provide much structure - at all - why would it be counterintuitive? If the goal is to let the DM figure out what makes sense - why would it be confusing?
And if the DM does a good job, why would it be unsatisfying?
It's not in-line with 5Es generally user-friendly approach and seems like a weird and unasked-for compromise, and also relies on bizarre Crawfordian takes on "natural language". In the end as you point out DMs frequently just overrule, house-rule or ignore it.
The guts of the surprise rules:
The DM determines who might be surprised. If neither side tries to be stealthy, they automatically notice each other. Otherwise, the DM compares the Dexterity (Stealth) checks of anyone hiding with the passive Wisdom (Perception) score of each creature on the opposing side. Any character or monster that doesn't notice a threat is surprised at the start of the encounter.
Is your hubbub about them ignoring the passive perception side of things? It doesn't make that much of a difference if you have the other side roll perception checks. It is just a bit of a dead giveaway to players that something is stealthing around them.
Beyond that, I'm at a loss as to what your concern is here. If noboy is being stealthy, then there is no surprise. If one or both sides are, then you roll and check versus passive perception and if your passive perception is not good enough to see any creature (stealthing or not), you're surprised. It just comes down to when the DM allows a stealth roll and when other combatants are far enough away as to not be noticed i there is no stealth roll from them.
So of course it's not changing for 2024!
It's fascinating how good BG3 is at accidentally highlighting issues with 5E's design, even as it makes wacky choices of its own.
I do not see this as BG3 highlighting the stealth rules for the RPG ... I see it as (outside of some advancements in AI) the game not being able to just have the DM decide and needing to have more precise rules because computers require more precision.[/quote]