Mentioned this to Pyro already, but a friend of mine's Hard ending level was L99. Make of that what you will.
For offensive item-use characters with a giant interp split on what they can use (Salve-Maker?), I suspect that I'd be happy with just leaving them out of the damage average entirely, and then never nominating them. For money throwers, I suspect I'll have to play through the game to tell. The key issue for me isn't amount of money thrown vs. money gotten per battle, but rather how much money is worth for other things at end-game and how much there is. If there's plenty of important stuff to buy until the final dungeon and you're generally cash-strapped, I don't particularly care that this eases up at the end, it isn't generally a viable move and the very endgame shouldn't be over-emphasized; if gold is plentiful and Not A Big Deal, then go nuts.
For allowable equipment, I'll have to get to the endgame to comment, but in general I suspect I'd be in favor of being pretty lenient with "uniqueness" for guaranteed treasure chest items but harsh on monster drops. I definitely reserve the right to change my mind, especially on power level concerns.
I implied this earlier, but I'd allow Default, and would certainly let the damage-reduction effect apply. I might apply a *light* defensive penalty to the entire cast if I'm feeling particularly fair (at absolute worst, it's +16% damage, but since Default -> 2x -> 4x attack isn't even that common a strat, it should probably be even lighter, maybe +5% damage), but this thing is minor enough that I'm fine with having it be a cast quirk. Basically there are three reasons to ban default: uniqueness, power level, and general DL philosophy. To ramble some:
1) Uniqueness - The fact that some classes have a functionally different Default command with bonuses dampens the uniqueness argument, and I in general don't care too much for this interp myself anyway; I usually allow cast-wide commands. This complaint is fair enough, though.
2) Power level, see Pyro's notes, it's actually not THAT big a deal. It matters in a straight slugfest only when psuedo-3HKOing, and only has a noticeable effect when psuedo-4HKOing or 5HKOing; if the BD job is giving the other side that much time, there's all sorts of nasty things that can happen (buffs, debuffs, low % status), and if the other side can't do 'em, they may well deserve to lose. Note that surgeable damage also tends to get around Default in a slugfest pretty well - do the one-shot of big damage on the non-Default turn. Default matters most for healers who can do the Default -> Heal + attack strat which basically requires the other side to do 67% durability damage to effectively 2HKO them; but healers tend to have weak damage, so they're even more affected by the high damage average. It won't be uncommon that healers have a psuedo-5-10HKO in the DL, so that's handing a LOT of time for criticals, charge-up moves, buffing, low % status, the works.
3) General DL philosophy, eh, not something I share, so I can't really talk about this, but I respect that some people just don't allow this to work! There's obviously some degenerate strategies that involve "guard" type moves depending on how you see DL matches proceeding sometimes, but I just don't consider them, problem solved. Certainly when uniqueness isn't the issue, there's all sorts of "charge up with damage reduction for big damage later" moves in the DL, and I presume most people let, say, Pokemon Fly's damage-nulling vs. non-Sky Uppercut type stuff work, so the idea of defensive buffs isn't something inherently bad for the DL, and this isn't even nulling.
