Perhaps taking a different tack would help explain my point of view better.
I'm fairly sure I understand your point of view. I just don't share it - or, rather, I see it as one viable way to play some RPGs, but not the only way. And not the way that I happen to be running my current, 4e, game.
RPGs benefit sometimes from having the players act on metagame knowledge.
<snip uncontentious examples>
Acting on metagame information can also be detrimental to RPGs. PCs are suddenly more cautious for no reason because the players failed some Listen checks.
<snip>
The former cases are praised, while the latter cases are scorned.
Here's where we part ways. The latter cases are not always scorned.
First, your example is ambiguous, between the PCs or the players becoming more cautious. In some games, there is no easy way to distinguish this - if the PCs are already fully bedecked with their armour and weapons, for example, and there are no buffs they can utilise, and the game has passive Perception mechanics, than there may be no mechanical, ingame way whereby the PCs can be more cautious. But if the players fail some Perception checks, and as a result start playing more cautiously - paying more attention to my descriptions, trying harder to puzzle out what excatly is going on, thinking harder about where some important thing or person might be hidden, etc - then I'm not necessarily going to scorn that at all! That might be part of the
point in getting them to roll a Perception check in the first place.
And, when it is the
GM who is playing an NPC/monster in a certain way based on metagame considerations (and that is what we are talking about in the context of fighters' marking ability), there is even less reason, in my view, to regard it as scorn-worthy.
when the fiction and the metagame mesh, the story tends to come out well and is aesthetically satisfying, as you said, whereas when they contradict, it doesn't make much sense and pulls people out of the story.
I'm not entirely sure what you mean by fiction and metagame "contradictin" - but I just don't regard it as true that metagame-driven fiction is necessarily unsatisfying.
When my players pull a "disguise ourselves magically and make the dwarf fighter, who has zero chance of pulling off a Bluff check, be the "prisoner" we are taking to the BBEG", the metagame allusion to Star Wars is obvious. But that doesn't pull us out of the story. It's part of the (corny) charm.
When the imp who pestered the party 3 levels ago suddenly turns up again as the familiar of a new NPC cleric enemy, in the course of the biggest battle the party has fought to date - defending a village against a squad of a dozen hobgoblins with a behemoth, plus bugbear "special forces", plus tiefling spellcasters, etc - the contrivance is, at a certain level, obvious. The fight is a "big thing", and the surpise attack from the invisible imp just adds to its "bigness". But, at least for my group, it doesn't pull them out of the story. It reinforces the story.
And then occasionally you get surprises. When they went hunting for the missing niece of the baron, whom they knew to be the fiance of a Vecna worshipper they'd killed, and after tracking her to an old necromancer's tower, I was sure that they would expect her to be at the top of the tower trying to revive a trapped vampire. But when they got to the top, and found the
niece as the key villain, they were genuinely surprised - they had been expecting to rescue her from necromancers. So sometimes the metagame contrivance, rather than producing an "of course" response, can catch them by surprise.
But these are all metagame contrivances. Admittedly these examples pertain more to encounter design than action resolution, but I think the considerations are much the same.
If the heroic fighter issues a challenge to all comers and vanquishes the dozens of orcs that come out to kill him, that's immersive; if the dozens of orcs attack him instead of the wizard or rogue because he's the fighter and he has more HP and AC than anyone else, that's not very immersive. That doesn't produce a narrative that you or I want to see.
Well, speak for yourself! The
point of marking mechanics - as I noted above - is to produce a fiction in which the fighter is at the centre of the action. And I have no trouble becoming absorbed by that fiction.
PC romance provides a different example. NPCs don't fall in love with PCs through any emotional process - as Hussar said, they are just constructs. NPCs fall in love with PCs because the players at the table decide that that is a satisfying way for the fiction to unfold. It's metagame-led fiction, but not, therefore, inherently unsatisfying.
The big difference between "good metagaming" and "bad metagaming" is that good metagaming can be (and usually is) justified in-game
<snip>
Yes, I'm a paragon of goodness, but I'll attack you just because you look...thief-y" and "So, apparently they have a counter for the super-secret spell I researched under tons of wards of secrecy; imagine that!" are things that are bad for the narrative.
The reason the things you describe as bad are bad - to the extent that they are - is, in my view, because they make for crappy play. The metagaming is neither here nor there.
If you take the primary feature of the fighter, the feature that's supposed to make him the rough-and-tough hero instead of the guy the monsters ignore on their way to gank the mage, and you make it a metagame thing and don't even bother paying lip service to the in-game rationale, that's bad metagaming.
At the risk of repeating myself, I disagree.
Here's another example from 4e. Paladins have a power called Valiant Strike. It gives a +1 to hit per adjacent foe. Why is it called Valiant Strike? Because it makes it more likely that the paladin (in order to get the bonus) will hurl him/herself into groups of foes. The power, due to its metagame consequences, will bring it about that the paladin is played as valiant.
I think that's clever game design. You may well disagree - and fair enough, not everyone like metagame-heavy mechanics - but to say that it is bad metagaming is just to beg the question in favour of your preferences and against mine.
It's fine to have marking be a purely in-game things. It's fine for it to be an in-game thing some of the time, and some of the time have it be an out-of-game thing because them's the rules, what are you gonna do? But to advocate that it doesn't matter what (if anything) marking means in game because you can just treat it as a metagame thing means that you're hurting the narrative.
As I said in my previous post, I have many actual play threads on these boards. I'm curious to learn where my casualness about the metagame operation of marking is hurting the narrative! I haven't noticed it.