It's already been mentioned that the golf bag encouraged players to learn/remember/know how different weapons worked. By y extension that improved the game in that it required players to put in some thought beyond 5e's "I showed up [and swing my magic weapon]". That also had the secondary benefit of allowing me the GM to provide a wider range of treasure for players to get excited about other than 5e's objectively better in every way or bust.Re: Focus Fire -- As others have pointed out, it makes a lot of sense to take down one enemy and thus remove someone attacking your side, and it is generally a reasonable course of action unless the other enemies can use their not-being-targeted to exceptional effect (this later part being the unrealistic part, as IRL those other combatants will be circling around you or attacking things you are defending or otherwise do need consideration, if nothing else some shots in their direction keeping them hunkered behind cover). Unfortunately, most of those things you can do to make one not want to ignore some enemies to finish off others are hard to model or draw out combat.
Others have mentioned previous editions where you couldn't necessarily choose your target for ranged attacks. I think another way of doing this would be to just have a greater difference in to-hit chance vs enemies 'in the front' vs. 'in the rear.' This would mean that if you wanted to focus-fire on the enemy glass cannon on an injured fighter, you would have significantly less likelihood of effect compared to taking on one of the front-liners the other side wants you to engage. This, to my mind (so, haven't playtested or anything) might go a long way to bring back some more varied and difficult tactical decisions without grievously effecting combat lengths.
There's a fundamental issue/tension with fantasy settings ('like medieval times, but with real magic and monsters') where you want knights and horses and swords and castles (or whatever trappings of the medieval world drew you to the setting in the first place), but those might not make the most sense in a world with actual magic and monsters. Do you put domes over your castles, since open-sky courtyards are defeated by flying enemies? Do your troops fight in formation, since that makes sense IRL, or spread out to avoid AOE spells? Do you fight giants (or dragons, or iron golems) with swords and spears or with qwertys and asdfgs (what are qwertys and asdfgs? why the weapons that would have been developed in a world full of dragons and giants)?
Back when there weren't any combat maneuvers and the combat was most abstract, I definitely did imagine that the fighters were climbing on the giants or leap-attacking or running under their legs and stabbing them in the butt when they squatted to smash them (my mental image of giants at the time were more 25-40', not sure what the rules said at the time).
Right tool for the job is an interesting issue in D&D. Specific Weapon vs. Specific Armor makes all the sense in the world for Chainmail, but I get why it didn't see much play in oD&D/AD&D. 'Can't even hurt' without silver or magic works to make some monsters scary, but often can reduce people to just their handy +2 silver hammer, since it always works. 3.5's different resistances worked, but that level tended to either incentivize a golf bag of +1-specificalignment-specificmaterial-specificB/P/S, or just saying screw it and powering through the resistance with high damage. It's a balancing act in trying to get the preferred playstyle to be incentivized.
If bob's +3 bow is going to mow through the mooks but to nothing against the big guy unless he switches to his +0 flaming bow it introduced some things a player needs to think about when a fight starts.