I'm all for cleverness from my players. I've probably already regaled you with the tale of the obsidian golem my players defeated by completely outsmarting me.
My issue is that I have seen--have at times
personally done--the wishing away of interesting challenges with a spell or two. Murder mysteries eliminated with
raise dead (thankfully, the spell Raise Dead in DW works differently, and is thus much more compatible with a murder mystery even when you roll a full success.) Exploration challenges obviated with a teleport, the classic "fly the Ring to Mordor on the Eagles" problem. I mean, 5e straight-up nerfed pretty much all charm-type spells
specifically because they were a notorious short-circuit for social stuff.
Again, I don't mind using tools creatively. I just think that, because of D&D's wild-and-wooly early design choices getting ossified as the Eternal Doctrine of Now and Forever, we're saddled with several low-level spells that completely negate a few specific gameplay challenges that would be interesting to solve with more effortful tools/approaches, and a few high-level spells that outright do no-sale entire plotlines. Some simple divination magic + a teleport or two turns an epic struggle into an afternoon's work, and there's...really not much you can do about that other than to take those spells away, whether outright or via nerfing them into something more workable.
In effect, it's the kryptonite problem. In the absence of kryptonite, Superman is invincible. In its presence, Superman is even weaker than an ordinary human. But unlike Superman, who can be challenged in ways that are totally orthogonal to his powers, spells can do literally almost anything, and that all too often means negating entire plotlines by, for example:
- Reading a target's mind
- Scrying a location and teleporting (as noted)
- Disintegrating a dangerous object immediately
- Reviving the dead
- Altering reality, even if under limits
- Creating indefinitely-stable backup clone bodies of VIPs
- Creating pocket planes to hide things in
- Nullifying all mental manipulation (and scrying, to boot)
- Traversing interplanar distances instantly
- Calling in outsiders to aid
Creating situations that still permit these things to be
useful, but which don't make them instant "I win" buttons for large swathes of meaningful conflict, becomes harder and harder. Death, containment, logistics, transportation, communication, surveillance, intrigue, and
most other forms of meaningful challenge go out the window, often forcing things into either a DM/player arms race, or rather repetitious stuff that can only be solved with brute force.
It's not a bad thing for players to get
some access to powerful tools that can be used in a lot of ways, some of them very creative. It
is a problem when those powerful tools become commonplace and largely suck out any potential for meaningful conflict and stakes that aren't "slugging match." When nearly all information is available to them with just a couple spells, and nearly all locations are reachable in a couple more, and nearly all interactions with others can be forced into the shape the players desire them to be...what exactly is
left? And they can do that in, at most, the span of 2-3 days. Less, if you have multiple spellcasters cooperating with one another, e.g. one each Bard, Cleric, and Wizard.
This is just one of several reasons why having a genuinely separate Rituals system is so useful. You can have these incredibly powerful tools, and even have those tools find their way into the players' hands--but those Rituals cost resources, some of which may be difficult or even impossible to replace. Being able to scry-and-fry on demand is a problem. Being able to do it
once, with planning and preparation, both enables and invites creativity rather than lazy constant use.
As stated: Adversarial DMing is an unfortunate baked-in element of classic D&D design, and as a result, we live in its frustratingly long shadow.