On the original topic: I think there may be a couple of staw men flailing away at eachother, here. Yes, in any edition of the game a 'fighter wall' that simply blocked access to the rest of the party (single fighter in a doorway, pair of them in a 10x10 corridor) was doing its job in protecting the rest of the party. And, yes, in classic D&D it was about the only way for a fighter to do his 'job' - and, classic D&D was all about the dungeon crawl so it was prettymuch SOP, and, at least IMEX, we got so used to monsters beating on the fighter we'd often have that happen even when there wasn't a choke point necessitating it. Even in the open field, the DM isn't far off base having animal-intelligence or enraged monsters attack whoever is nearest or hit them in melee last or whatever (that's not so unrealistic - or 'wolf pack tactics' wouldn't work for wolves IRL).
3e gave us AoOs, and the fighter's high BAB made him good at them. A feat or two, and a fighter could take several per round, trip with them, and threaten at reach. That let him hold a choke 'point' 25' across. Add a spiked chain and an enlarge potion and he could form a one-man redoubt exerting control vs enemy movement in a 50' diameter circle, potentially protecting allies huddled close to him even in the most open of open fields. Sure, there were plenty of things that rendered the fighter irrelevant in 3.5, but it wasn't impossible for a fighter to exert some very meaningful battlefield control, if you exerted your system mastery (or consulted CharOp) and built to that goal.
4e actually took away threatening reach from pole-arms and the spiked chain, and thus the fighter, but, with combat superiority and per /turn/ OAs, every fighter could hold a 15' choke 'point' pretty well, marking those who did slip by as a bonus. Marking also made the fighter (and every fighter, not just specialized builds) more consistently relevant in an open-field fight, too. But the fighter only marks what he attacks, and his close (AE) attacks are all encounters or dailies, so his ability to mark a whole mass of enemies is limited (and under the players control). And, that's not a bad thing since the fighter's AC & HPs aren't /that/ superior to other characters and getting surrounded and beaten down is still very much a bad thing - so holding that choke point remains, "realistically" a very good tactic.
What pulled 4e away from the 'doorway' fight more than fighter stickiness was increased mobility (due to doing away with the Full Attack and Full Round casting times), and the greater viability of melee combatants. There are a lot of very effective and fun melee classes in 4e, and even classes that are primarily ranged can handle melee or have powers or other tricks that work well in melee or at very close range. So, there can be an incentive to get into the room and have several party members fully participating in melee, rather than casters and archers standing back from the fighter in the doorway and making the room beyond into a beaten zone (but, if there is, it's the result of players wanting to melee in the first place). When outnumbered finding (or having the controller create) a choke-point and defending it is still a great idea, but it's not the only idea. And, depending on class mix, builds and play-style, you could still have a very traditional fighter wall or a fighter at a choke point guarding a wizard and several ranged strikers who do the real killing.
The 5e playtest may hint at a return to cater dominance, but it has three melee-capable characters out of 5, and no hint of a return to full attacks or full-round casting times. So mobility remains and the 'doorway' shouldn't be too big a 'problem' or still a strong tactical option, depending on how you look at it. Open field battles, as it stands, could turn comical, with everything running past the fighter to attack the squishies, the squishies running away and casting, and the fighter running after the monsters (que Benny Hill music), but at least no one's losing 'full attacks' to all that running around, and even a little bit of stickiness - marks or tripping or grabbing or OAs for disengaging (1e-style 'parting shots') - could moderate that.
On the "Realism" issue: Even granting that realism is desirable, it needs to be applied even-handedly and without hurting the game, as a game.
If you want a realistic world that uses medieval weapons & armor and tactics, then the impact of magic on that world will have to be pretty limited, or tactics will have changed. Just making magic 'rare' doesn't help, either, because that only makes magic-wielding characters /more/ powerful as enemies are consistently un-prepared for what they can do.
If you want a level of realism among characters, you have to bring everything, including magic, down to that level so the characters balanced and the game playable and fun for all at that level of realism. It may seem, ironically, 'unrealistic' to peg magic to any level of realism, since there's no RL magic to serve as a model, but, for that same reason, magic can be given whatever level of power and/or limitations needed to balanced it, because there is no 'realistic' standard it has to match - magic can always be defined arbitrarily.