EDIT: I'll wager that, for some fans of low-op play, the inclusion of any of those classes excluded above in this post would limit viable character concepts, requiring justification as well. Consider it the "I'd like to play a ranger, but why would a party of a druid, a conjurer, and a shaper psion need a ranger tagging along?" concern, expressed by the ranger player, and not the high-op guys.
Ranger is a concept as well as a class. I can make a character who does all the things a Ranger
should be able to do -- meaning all the things it says that does but doesn't happen to be able to do -- who can totally keep up with, if not exceed, your Druids, Conjurers, Summoners, etc. At least for any practical level of optimization (i.e., setting aside TO type stuff or the most ridiculously optimized builds).
So, the limitation you point out is chimerical.
Now, is it possible to play a classic D&D Wizard in a low OP way? Can you play someone who summons things or casts Solid Fog and still be low OP? If so, then I'll withdraw the objection. Pathfinder's Summoner might be an attempt that sort of thing, maybe?
That being said, I think there is a cost that only goes in one direction. Telling someone they can't play the Ranger class -- note the class,
not the concept -- b/c that class fails at all the things it says it's supposed to be good at. That is, the player will not be the Aragorn expy he wants, but instead will fail at all the things he's good at, comes at very little cost. You're sparing them sadness and disappointment. Telling a player there's a whole set of concepts off-limits, that no, he can't play that Raistlin expy he desires, or he can't play that Cleric of Kossuth, is much worse/more costly.
Again, if there's a low OP option, even just a way of building a low OP Wizard, then that problem is a lot less severe. And, I will freely admit that the lack of an obvious one may be due to a certian breed of myopia or selection effect -- this is a charopp forum after all.
P.S.: I think there's a lot of slippage between Tier and OP. I could be mistaken, but I believe the above post is equating them, reading low-OP as low-Tier, and high-OP as high-Tier. I don't think that's the right approach, though. If I play the most OP Factotum or Duskblade or Warblade I can make, (or more controversially but possibly true, Rogue or Monk) it's still going to be "high OP."
P.P.S.: on a personal note, I had only given this a little bit of thought prior to participating in this thread. But, I don't have much against a low OP campaign, and would be happy to participate in one. My only requirements are that it'd have to be made clear at the outset -- that way I don't get invested in a character or concept that then gets determined to be "off limits." I'd probably also prefer the optimization to be totally determined by the Tiers system, which while it has its uses I don't think is necessarily the best measure of these things.