Hey, thanks for the responses, everyone. The problems are starting to come into focus.
Lets see if I'm understanding you.
1. Auto success or auto fail:
In terms of battle resolution, it looks like some monster stat tweaks may be needed in 3.5, but it also seems that having the party fight multiple monsters--as in 4e--would help. If an encounter has a high damage/ high hp/ low AC enemy, a high AC/ low hp/ low damage enemy, one that flies, one that has to get up close and personal, etc., that seems like that would give both high BAB and low BAB folks, and the melee and ranged folks something to do (successfully) during the fight. True?
Also, I like that 4e added Dex to ranged damage, so perhaps use the same mechanic for spells: Add Int mod to attacks and damage from Intelligence based spells, (or somehow bump up the area of effect, or duration?), and Wis mod to Wisdom based. That might help the low BAB folks connect with higher level targets.
I know, I know--thats more math. Ok forget it.
In terms of saves, again, 4e might have this right with adding 1/2 level. But how to reduce the spread between best and worst saves? Completely redo the save progression table? Allow more magical enhancements to just one type of save, (to balance high and low)? Or again, do monster and spell stats need to be adjusted in regard to save vs.?
2. Too many choices slow down combat:
Modifying the 4e powers template for 3.5 spells and attacks would help--everything laid out clearly and no flipping through books!
I also make it a point to list to hit and damage for each weapon, and also for each weapon or spell when buffed and/or when using different feats. At high levels, its a long list, but I'd rather scroll down to see "Greatsword, PA 5, Bull's Strength", followed by the appropriate attack and damage mods than have to do the math every time. Which brings us to:
3. Too much math:
I think buffs, level drain, conditions and all the other modifiers that can affect ability scores and attacks make 3.5 combat fun. There has to be a way to keep all that, make it all easier to track and make the math easier to figure on the fly.
Maybe a limit on the number of modifiers on any one target? Some sort of digital tool would be great--just select the target, then add the effect from a drop-down, and all abilities and modifiers are adjusted.--like how Kloogewerks does it.
4. DM prep time:
Templates for NPCs, or hell, just a big book of pre-rolled NPCs of different types and challenge levels sounds like it would help. And pre-packaging the same monsters for different challenge levels should also help speed things up. But is the time suck more about "What would be a good challenge for this party?"--walking that thin line between "just challenging enough" and TPK?
Am I in the ballpark with any of this?