Having played Pathfinder lately, concentration is a godsend for preventing buff stacking. I had many end game fights where, before rushing in to face a boss, the party would stop and cast buff spells for a solid minute.
And every NPC in Pathfinder tended to have variant statblocks of them without buffs as most had a half-dozen magical effects on them. <...> Plus, knowing that you can knock out someone's concentration and end a negative spell affecting an ally is great and allows for some amazing teamwork. Dropping the evil wizard's concentration has led to some great moments at my table.
As I said previously "I don't mind either in principle, but the execution of both just throws me out of the fiction."
I am 100% with you that Pathfinder, like its 3.5 daddy, is a nightmare of buffs and stacking rules. I
totally agree these needed to be knocked back, maybe not as much as Lanefan would like (i.e., to none) but a fair bit. Advantage is a great example of how one can avoid painful stacking rules in a simple way. Temporary Hit Points is another example.
However, WotC doesn't have much of a clear rationale for one spell being concentration versus another much of the time. I often think they just kind of fly by the seats of their pants with the decision. Consider
Mirror Image vs.
Blur. The spells were different historically because the former was a Magic User spell (for some reason) and the latter was an Illusionist spell back in 1E. The spells are functionally similar---making it harder for the caster to be hit---and probably should just be the same spell.
Mirror Image has a fiddly and annoying mechanic and no concentration.
Blur has a really clean mechanic (disadvantage on attacks) but requires concentration. My guess is that many people don't take
Blur just to avoid concentration even though as DM I'm pretty sure I'd much rather deal with
Blur than
Mirror Image. I know I go out of my way to avoid spells with it just so I don't have to keep track of it. A way to rewrite the spell to make it clean without concentration might be something like:
Level 2 Blurry Image. Duration 1 minute. This spell induces a shifting, blurry mass where you are standing, making it hard for attackers to hit you. The next three attack rolls against you by any attacker without Truesight or Blindsight are made with disadvantage. Cast at a higher level: For every additional level slot used to cast this spell, add two more attack rolls made with disadvantage.
Then the fact that there's no method for eliminating it, for example by casting with a higher slot. This means there are spells that are essentially "don't bother" because they require Concentration.
Web is a very good example of this. At level 3 it's cool but it rapidly becomes useless at higher levels for anyone but perhaps a wizard so it's pretty unlikely a character with a limited pool of spells would take it.
It needs to be a number, so why not three? Three has
literary significance and religious significance.
So? Sounds like a Jeremy Crawford ad hoc rationalization. Those significances aren't in the fiction, even remotely.
Three is also just the baseline. If you need more because you're handing out more magic items and want a more monty haul game, then it becomes four or five. If you have a magic-lite setting with that kind of item being rare and special, then perhaps a single attunement slot works.
It is, though systems like D&D Beyond enforce it and disallow any violations of the rules, which is certainly a pain for folks who use it and might want to enact some kind of variation for their dreaded Monty Haul campaign. (I don't know how other online systems work, so I can't comment on them.)
Furthermore, once again like concentration WotC just uses it as a balancer of sorts when they think an item is maybe a bit too potent but without some kind of clear rationale. For example, there are items that one might get at low levels that are, like Web, fairly cool then. They may be interesting and thematic but don't scale and are essentially blocks in the way of a more useful attunement slot. The
Ring of Mind Shielding is a good example. I can see why it requires attunement, too but it just seems to present players with a lot of rolling build traps. One thing I noticed over the course of a long campaign (that went to level 20) was that many items simply provoked a "meh, I'm not going to bother" reaction from players because they were stuck with the tradeoffs of which of their old items to eliminate. Much like with buffs, I totally get the reason for a limit of some sort but it would be nice if it scaled somehow.
In sum, disliking the implementation of something is not saying that it serves no purpose. Yes, I could devise something like that myself but working out the bugs of a system like that is why I pay game designers!