Two points here.
I like the Come and Get It Power, a lot. I love the way it simulates the scene in which the tough ol'fighter goads a load of mooks into charging him. Sure, there are other ways to make that happen, but they are either inelegant, or require the DM to agree with the player. The power solution is great.
"Narrative" facets of D&D work well, and satisfy the "traditional" crowd, in two cases. First, when they have an in-game rationale provided which matches the mechanics. For instance, the Goad feat in 3e forces targets who threaten you to attack you in melee instead of someone else, and requires a Will save against a Cha-based DC, affects only intelligent creatures, and is mind-affecting. That means that while players can (and often do) object to the way it works ("martial mind control!") they can still take those points and determine what's actually happening in-game (it's Cha vs. Will, so it's like intimidate, it affects intelligent creatures only, so it's based on language or the like rather than body language that would work on animals, etc.) and there are defined ways to defend against it (mind-affecting immunity) that make sense in-game.
Goad is basically "martial mind control," but it is
explicitly so, as it has the [mind-affecting] tag, so anything that would logically protect against mind control works against that too. So while it doesn't simulate
our reality at all, it is internally self-consistent with D&D physics and action movie physics, which gives players a point of reference to start using it/changing it/whatever from. Come and Get It does not provide such an in-game rationale. It happens automatically regardless of enemy type, it forces movement in a particular direction rather than making their attack choice for them, and doesn't have any way to protect against it. Even something simple like an attack of Cha vs. Will or even Str vs. Will would reduce the "automatic mind control" aspect and convey the fact that he's doing something scary/taunting to intimidate opponents, but the fact that it just works is a major part of the problem. It's good that a fighter can use it to draw out mooks action-movie style, it's bad that he can also draw out bosses, mindless skeletons, dishonorable villains, and any other things too.
The second case is a purely metagame mechanic, when the mechanics are invisible to characters. Action points are a good example, because while the player can use them to affect the narrative the character has no idea how or when an action point is used or even what one is beyond the fact that they're luckier than normal when it counts. So "traditional" gamers are not averse to narrative mechanics, they're averse to ones that affect the narrative unintuitively or make narrative tropes available/visible to the characters when they shouldn't be.
A DM can easily rule that it doesn't work on a given creature, too. Fourth edition marked a return to sensible DM rulings, something that 3e stepped far away from.
Let's keep that baby, shall we? DMs make rulings. There isn't, nor should there be, a rule for everything.
Another rant, once again spoilered for length:
[sblock]See my earlier rant regarding page 42 and the gap between codified rules and ad-hoc rulings. You may have heard of the Oberoni Fallacy before, which says that "The DM can fix it, so it's not broken" is not valid reasoning. Yes, the DM
can rule that you can't trip an ooze, the DM
can rule that CaGI doesn't work on certain opponents, the DM
can work to make the 15 minute day not happen,
but he should not have to. 3e and 4e are both guilty of this in some respects, but it's slightly more problematic in 4e due to how much more they've codified everything else.
You're right that there shouldn't have to be a rule for everything, but there are two caveats. First, you can't say "I don't want to have to ask my DM if I can do this! I'm glad it works exactly as it says it does without the DM having to agree with the player!" and then say "Well, you know, the DM can change stuff if he wants, there shouldn't be a rule for everything...." Either codify everything to the same standards as everything else, or leave a hole in the rules for the DM to fill as desired, but don't try to do both with the same thing at the same time--and if you're going to do both for different aspects of the game, make things that are the same in every game (character-based stuff like powers and classes) codified and things that will differ between games (setting-based stuff like monster frequency and item availability) open to rulings.
Second, the more frequently you have to make rulings, the worse the game will play. And let me clarify that before you say "But AD&D--!" In AD&D (which I played and DMed for years, by the way, I'm not new as of 3e) there are certain areas of the rules that are left for DMs to fill in and there are certain areas that a DM might want to change. If you're playing 1e and you find that it doesn't have a bartering system, or you find that you want to alter how weapon specialization works, you can make a ruling and move on. Once you make that ruling, it's set that way for your games, and everyone knows how it works moving forward. However, those kinds of rulings are what we call houserules nowadays--AD&D DMs may have had pages and pages of houserules in big binders to be explained to new players, but once those new rules were explained they didn't have to be again.
The more you have to make rulings during the game, the more things will slow down, because of the cascading effects, increased interruption of play, and the increased complexity of newer rulesets. If a 1e fighter wants to trip an ooze, has asks the DM, the DM can just say "no," the game moves on. If a 4e fighter wants to use a power on an enemy and the DM says no, (A) that affects that power from now on, (B) it has implications for similar powers, (C) he has to say "no, actually, that didn't do X, it only does Y" and cause takebacks and discussions and such because the default assumption is that the power works as-written, and (D) 4e combats involve many more discrete effects per character and many more characters, vastly increasing the opportunities for rulings and slowdowns when rulings are made.
The variation in effects in the power system makes this worse, as every exception is a point of failure to introduce new arguments and rulings. If a 3e DM decides that you can't push an ooze without magic, he makes a ruling on the bull rush combat maneuver and moves on, because basically everything involving moving enemies uses a function call to bull rush and there's only one special case (nonmagic bull rush vs magic bull rush). "Bull rush doesn't work on Oozes without magic" is a nice simple rule; bull rush is defined, Ooze is defined, and (Ex) vs. (Sp/Su) is defined. If a 4e DM decides you can't move an ooze, he has to make rulings on different powers individually because (A) the powers have different mechanisms (persuasion vs. brute force vs. etc.) and (B) there's no "ooze" category to apply changes to, so he'd have to determine this for individual monsters. "Forced movement that relies on physical force cannot move an amorphous creature" is not a good rule, as "physical forced movement" and "amorphous creature" are not tags, therefore leading to corner-cases where things perhaps should work but don't or vice versa.
Ideally, a game should be able to run without a DM contradicting anything--there may be areas where the rules say "DMs, make something up," and those should be few and far between in modern games, but the game should be able to play well and cohere without the DM having to step in and rewrite the grapple and diplomacy rules from scratch or make rulings on a bunch of powers. Once the game is functional, then the DM can make whatever houserules and rulings he wants. Houserules, rulings, and custom contents are there to make a game work
better for an individual group, not to make the game work
at all, because DM quality varies drastically among groups as I'm sure anyone who has played under bad DMs is aware and because rules-heavy systems relying on rulings to work or encouraging designers to lazily rely on DMs to fix things is frankly embarrassing today.
I'm happy to make on-the-fly rulings when I run AD&D, as the system is fairly rules-light and flexible relative to modern games and many things resolve faster. When I run 3e and 4e, I keep rulings to a minimum and rely on houserules, letting everything else play out as the rules indicate. Rulings are not inherently bad, I hope that's not the message people take away from this rant, but different systems work better with rulings vs. houserules vs. rules and different areas are better to rule than others. For 5e to rely on rulings would be an absolutely terrible decision if it's based on 3e or 4e (we know better with 3e and codification is much higher with 4e) and still bad but not as much if it's based on AD&D.[/sblock]