There are three major problems that I see with the disintegrate nerf.
1. It is now conceptually incoherent.
The spell can disintegrate a 10' cube of solid stone, but can't destroy a human being, even on a failed save? That seems slightly bizarre. Hong rebutted this when I posted it on another thread by asserting that since high-level characters are 'special', they should survive- after all, Joe Commoner is going down to disintegrate. Fine. How about an elephant? With 104 hit points, it has a good chance of surviving having a 10' cube chunk taken out of it. How does that figure?
2. It is now mechanically inconsistent.
I could understand, but disagree, if they nerfed all save-or-dies. I note that the discussion has turned into the old 'save-or-die: yea or nay?' argument of multiple threads past. The point is that they nerfed disintegrate, but not slay living; nor finger of death; nor wail of the banshee. Very few of the anti-save-or-dies here favour any save-or-die spells: just look at the arguments being used.
As for Disintegrate being stronger because it bypasses constructs, undead and Death Ward- so what? Constructs are usually immune to all magic (unless you're going toe-to-toe with paragon homunculi on a regular basis). Undead, at high level, are reasonably likely to be incorporeal. Death Ward is far less common than spells which boost touch AC- given the duration, it is unlikely to see it against a non-prepared enemy. It is far less likely to be seen that high-level touch AC boosting spells and items: Rings of Deflection are reasonably standard by high levels, and everyone should have a Dex 13 or better. Cover and concealment are both far more effective against DIsintegrate than FoD.
3. It's weaker than direct-damage spells.
Chain Lightning, the archetypal level 6 damage spell, caps at 20d6. Assuming that Disintegrate is 20d6 + d6/level, it caps at 40d6 in non-epic play. CL has a save for 10d6, Disintegrate for 5d6. We shall assume saves are made 50/50 (WotC guidelines are that good saves are made 2/3; bad ones are made 1/3. Empowered^X Fox's Cunning is no longer an issue, Spell Focus is nerfed, and in any case good Fort saves are on balance more common than good Ref saves). Given these parameters, the average damage is 22.5d6 for Disintegrate; 15d6 for Chain Lightning. However, given that Disintegrate has make a touch attack, if the chance to miss is 35% or more (i.e. needing just a 8 or better to hit), then CL is superior- not counting concealment or cover. If, as some have asserted, Cloaks of Displacement are so common at high levels, this hurts Disintegrate immensely whilst leaving the DD spells totally untouched. Moreover, classes with poor touch ACs tend to have good Fort saves (dragons, fighters, clerics, giants) whilst those with poor Fort saves typically have good ACs (rogues, arcane casters, undead). Furthermore, CL affects secondary targets, and has a far greater range. On balance, you'd almost certainly be better off with CL than Disintegrate, unless you have an absurd ranged touch attack, min/maxed DCs or are fighting creatures with poor touch ACs *and* poor Fort saves (i.e. nearly no-one). Incidentally, CL is often regarded as slightly too weak at initial levels, compared with Maximised Fireball. And it's patently inferior to Empowered Flame Strike.
The new Disintegrate is conceptually incoherent and weak compared with either traditional save-or-dies and direct-damage spells. There are conditions under which one would take it: multiclassed rog/wiz for sneak attack potential, wizards with very very good ranged touch attacks, wizards with DCs in the stratosphere (much more difficult in 3.5e), when facing opponents with poor Fort and poor touch ACs (very few), or as a general utility spell. However, in most situtations, it's outclassed by other spells at a comparable level- or lower (Slay Living, 5th).
