Viva la 3.0E!
I
would be playing 3.0 still, if I could find anyone else running it. :-/ But it seems like all the gamers I know apparently got rid of their 3.0 books after switching to 3.5, so I've been outta luck. I still have my 3.0 books (and very little of 3.5), and I've still tried to run 3.0 campaigns, but it's hard to find players (let alone ones who stick around more than a few sessions) for an out-of-print game that went through that kind of quick revision, for the reason already noted.
I ran a 3.0 campaign for about 2 years or so, a few years ago, and some shorter campaigns before that. It runs a bit more smoothly than 3.5 in some regards (no conflicting rules from different sources except for a minor revision of the Polymorph Self/Other text in Tome & Blood and later Masters of the Wild, Conjurations actually allowed Spell Resistance in many cases rather than being wierd exceptions for no good reason, golems' Magic Immunity was simpler, the Orb spells in Tome & Blood were Evocations and actually balanced or just a little subpar at worst, XP was determined based on APL and the challenge overcome rather than individual calculations for every PC of different levels, monsters tended to have broader climate/terrain options instead of being shoehorned into one type of territory,
playable races weren't quite so far apart in power, gnomes weren't shoehorned into some bardic role, half-elves weren't shoehorned into being diplomats, standing up didn't provoke AoOs so no infinite-trip combo, Damage Reduction was much simpler and didn't require golfbags full of different weapons, weapon sizes were nice and simple and actually made sense, magic loot was easier to divvy up and use, the Alchemy skill didn't require being a spellcaster, Animal Empathy was a skill rather than a class feature, animal companions were normal animals and you could have a bunch of 'em instead of just one magically-augmented quasi-animal, prestige classes were clearly limited by the campaign and DM's judgment and they were clearly marked out as not incurring multiclass penalties,
clerics and druids didn't outclass everyone else quite so much, there wasn't much class/race/monster/feat/spell/item bloat, paladins were actually worth playing at low levels, Improved Critical and Keen stacked like they were supposed to so you could actually use weapons other than longswords/greatswords without sucking, Power Attack was simple and reasonable so it was okay to use other fighting styles besides two-handed weapon Power Attacking, magic projectile weapons stacked their enhancement bonuses with those of their ammunition so archers and such weren't less effective than melee fighters and there was actually value in buying or finding magic ammo,
Darkness spells
actually created total darkness instead of wierd shadowy illumination, Ray of Enfeeblement didn't allow a saving throw so it was actually useful yet not overpowered, Enlarge and Reduce were simple and unproblematic, Bull's Strength etc. actually had good durations but variable bonuses so they weren't broken, Divine Power and Righteous Might were simple and relatively fair, Emotion and Symbol and other spells actually had options instead of every spell being bland and single-purpose, sorcerers and bards weren't as screwed in spell selection though they also didn't get to swap out old spells, Deathwatch wasn't inexplicably Evil, Transmutation was a diverse spell school instead of a gimmicky support school, and it was easier to keep track of the rules and improvise when desired without the rules getting in the way as much)...... I could probably go on, but I think I've made my rambling point.
There were certainly a few flaws, but they were generally easy to spot and houserule into a fixed state. Like Harm/Heal, just houserule them to allow a Will save for half damage when appropriate (like Cure/Inflict spells). Haste is a strong spell and favors spellcasters in 3.0, but only affects one subject in 3.0; those who consider it problematic can just houserule away the extra partial action it grants, or limit its partial action to mundane actions like attacking or extraordinary ability use if they really feel the need to; I don't.
If your players are really whiny about PC-death, you might houserule Raise Dead to have a casting time of just 1 round to make it somewhat less onerous (while still leaving the higher-level, more-effective rezzes with their longer casting times). And maybe reduce its material component cost.
The Toughness feat should probably be houseruled to give more HP over time, like +1 HP/3 HD besides the 3 HP it already gives initially. Ambidexterity could possibly be interpreted in a more-liberal and more-beneficial way as allowing off-hand attacks to use the full Strength bonus to damage, which would make two-weapon fighting a more balanced fighting style choice (and would make Ambidexterity not suck).
Not much really needed fixing in the 3.0 core rules, and it was generally just minor stuff. A few prestige classes (Frenzied Berserker, Forsaker, Shifter, etc.) needed some fixing (which 3.5 didn't really do anyway), but as long as you actually enforce the prestige class background/flavor-text requirements and DM-judgment restrictions, then they shouldn't be a problem (meek little midget Frenzied Berserker with mediocre Str/high Dex and an amusing personality? Sure, go right ahead!). And even Monkey Grip in Sword & Fist is quite fair (perhaps even a bit weak), unlike its broken 3.5 version, which is also true of some other 3.0 feats, spells, and classes.
3.5 tended to emphasize prestige classes as more of a powergaming optimization choice than a stylistic fun choice, though there were certainly some 3.0 PrCs that were notably overpowered (many full-casting-progression PrCs with significant extra class features, for example, and stuff like Frenzied Berserker; generally Masters of the Wild had some overly-strong PrCs and spells that gave druids too much wizardlike and clericlike versatility, even though the rest of its content was decent, and some of its spell/PrCs were decent as well). Loremasters and a few others are alright because of their limited, tight focus and significant prerequisites, while several 3.0 PrCs had only partial spellcasting progression (75% or 50% usually) or an otherwise-weak independant spellcasting progression, to make up for their significant extra features.
If running a 3.0 game, I'd tend to recommend using the core rules + the 5 guidebooks (Sword & Fist, Tome & Blood, Defenders of the Faith, Song & Silence, and Masters of the Wild, or a similar basic supplement-series from a different publisher) + Monster Manual II (or another monster supplement, like Fiend Folio or Monsters of Faerun or even Oriental Adventures), and maybe the Manual of the Planes, or Arms & Equipment Guide, or Stronghold Builder's Guidebook. But the core 3 + some class guidebooks is generally enough for a decent variety of options.
I've been considering starting another 3.0 campaign over OpenRPG sometime soon, anyway, if I can find enough players for it. Probably gonna use my Rhunaria setting/houserules or my expanded version of Rokugan (again, since I've already run 3.0 campaigns in those before), since I don't have my Aurelia setting/HR done yet. I really need to get a new website to host my Rhunaria and modified Rokugan stuff again, though (Geocities is no more and Yahoo! has changed their services too much).