I've got a lot of contention with many of your statements, mainly because you are jumping from point to point on the assumption that your points are universal axioms of the game, when in truth they are highly specific, totally subjective situations. To expand:
Advocates of this "rulings over rules" approach often overlook that such games are not games at all, but rather DM-story telling exercises (making it up as he goes along) with the players just watching for lack of true agency; so they don't really even need to be there! Sub systems with transparent/impartial mechanics can mitigate that DM-as-puppeteer problem by giving all participants defined mechanics for influencing play.
This is a prime example of a very specific approach to the game that shouldn't be used as a blanket statement for everyone else. "Rulings over rules" is used to get past a sticky bit which is slowing down play; it is not part of some greater storytelling agenda to rob players of their agency. It simply means that when the DM is getting real damn tired of his players having to look up the grapple rules every time they try to grapple, he can just say "roll a die; if it beats X, then he's grappled. We are always doing it like that from now on because I am tired of grapples slowing things down". That's all. It does not mean hand-waving of all the rules so the DM can tell a story while the players sit there.
Subsystems should be considered for addition or omission to the game based on two factors: how much fun they add to the game, or how much tedium they cut out of the game. Whenever you have the choice to add or omit something, it should always be based on how much more fun it makes the game (for as much of the party as possible). If your goal is a simulationist game, then simulationist rules are likely to be a good addition because they expand the simulation. But if your game is not a simulationist game, then there's not much fun to be had in adding a rule the requires players to re-sole their boots once a month in order to avoid suffering a movement penalty or whatever.
Back to AC adjustments, it's NOT about "realism"; it's about being "plausible" and "historically accurate." Your reference to the footman's flail illustrates the strengths of AC adjustments well; they give players historically accurate incentives to use the full range of medieval weapons in plausible ways (the flail was prized for its ability to crush armor). The drawback of the footman's flail is that it requires both hands - no shield. AC adjustments can and should drive decisions prior to combat. If you know that you are likely to encounter a certain AC type in the journey ahead, then bring the right weapons to deal with it. Again, same logic as for spells. This in turn gives fighters incentive (and value) in carrying more than one weapon. It's a GOOD thing that the AC adjustments incentivize the use of weapons other than long swords. There are very real reasons why all those other weapons evolved and why by the end of the middle ages the halberd dominated; AC adjustments help bring that out in the game.
First I'd argue that the terms "realism" and "plausible/historically accurate" are synonymous, but that's just nitpicking.
I think Beoric's points about the weapons system is that it seems "unrefined"; that attacking with a weapon is a pretty damn common part of the game, and should therefore be much more streamlined than the table-consulting ways of old. The d20 system made this all very smooth; you can tell if a weapon hits or misses without consulting anything. You roll the die, add the number you always add, and hope you get higher than the AC. Very straightforward, very smooth, very fast.
Yes, some weapons will be better in some situations, just like in real life. But again, newer systems account for that, with Resistances/Vulnerabilities, Advantage/Disadvantage, as well as simplified weapon characteristics (like "Heavy" which simply means small creatures use it with Disadvantage; "Versatile" which means it causes a different damage if being used two-handed as opposed to one-handed; "Finesse" which means you apply your DEX modifier to hit rather than your STR modifier, etc.) Different weapons are made more or less desirable in a situation based on these qualities - the big difference is that they can be tracked in your head, saving reference time/table space. If you have a Heavy weapon, you already know what it entails - you don't need to refer to "Table 3A - Heavy Damage Variants" or whatever. It's just a way a weapon is, like how it's inherently known that bows need ammunition or that a hammer is a bludgeoning weapon.
You get the same considerations and benefits for carrying multiple weapons, but it's just more streamlined. In newer editions, you'll still get situations like "oh shit, skeletons ahead, better swap my sword for a mace", but it's based on skeletons having resistance to piercing & slashing weapons, not on having to look on a table to see that a mace has a +2 against a skeleton AC and a sword has -2, or whatever.
Without ACs adjustments, 1e weapon choice usually boils down to one of these "gamey" statements, all of which I've actually heard at the table. "I'll pick longsword because most magic weapons are swords and most swords are longswords. Why would I use a two-handed weapon that only does d6 damage when I can use a one-handed weapon (long sword) that does d8 damage? Let's bring a mace in case we meet skeletons."
Again, sounds very table-specific, which means it's not good to just declare that the system "should be this way" simply because it better suits your very specific situation. My players, for example, would never say "most magic weapons are longswords", because in my game most magic weapons aren't (I tend to put in more magic bows and daggers than anything else), so why would I ever need to consider that as a criteria to influence which system works better than another? Or as to why a character would use a lower-damage weapon compared to a higher damage one - maybe the player isn't trying to min/max damage output like some DPS WoW player, and just wants a weapon that fits the character thematically?
Thing is, it's not really an issue in newer editions because of the weapon trait system (having the Finesse characteristic, having the Light characteristic, etc.). It's only an issue with the table system, specifically because it's using an outdated format for figuring out different weapon advantages and disadvantages in specific situations. Having to calculate certain modifiers in situation-specific encounters is exactly the sort of bloat that put people off 3e.
