That's a campaign specific thing. My dungeons frequently had "traps" in the form of "check for disease exposure" (there are fleas on the dead animal, for example) and I did do monthly checks for disease exposure. And I did enjoy Wererats as foes quite a lot, and that's not just lycanthropy but lots of diseased rats. And so forth.
Agreed it's campaign-specific. I can't remember the last time disease played any significant role in my game; though in the game I play in the current adventure is all about disease and fighting the agents of a disease deity, and immunity would be really handy!
I didn't really hate immunities until 3e when the game began to become all about immunities, but in general 3e's attempts to move things away from immunity and more to relative immunity where they had considered it - like DR replacing "a magic weapon is needed to hit" - were things I found really welcome. Complete immunity to disease from 1st level is a massive advantage, as would be complete immunity to fear (did I say I use the Ravenloft fear rules in normal campaigns?).
Immunity to fear (a.k.a. "fearless") hasn't made Cavaliers any more popular round here, and they have the worst survival rate of any class (caveat: they also have a small sample size).
I was never a big fan of DR as it just tended to make long combats longer. With "needs magic to hit", if you've got magic you affect the thing normally and if you don't there's no combat - you run away instead.
Not really. Because before weapon specialization, fighter wasn't all that great. Paladin was just fighter but better, and Ranger was fighter with a lot of extra powers and better starting hit points and more panache. Specialization made the fighter on par with the Cleric and M-U and dragged it to competitive or better with UA classes like Barbarian and Cavalier. If you take out specialization, then the game actually gets less balanced unless you just stick with the core four classes.
I'm not saying take specialization out entirely, I'm saying tone it down. Have the benefits accrue slowly, something like
Level --- to-hit --- damage --- attack rate
1st --- +1 --- +0 --- 1/1
2nd --- +1 --- +1 --- 1/1
3rd --- +1 --- +2 --- 1/1
4th --- +1 --- +2 --- 3/2
7th --- +1 --- +2 --- 2/1
10th --- +1 --- +2 --- 5/2
By 4th-5th level they've usually got or are getting weapons that give decent plusses as well, and by 7th-ish level things like gauntlets and girdles of strength start showing up, so having the plusses from spec. cap after 4th isn't a big problem because the extra attacks are far more relevant.
UPDATE: Going a bit further though, the problem with specialization isn't so much that it effects party balance. Thieves get even worse, but they were bad to begin with. The problem with specialization is that it heavily effects balance between the party and monsters. The solution I found was that when the party got powerful enough, to make the monsters more powerful to compensate. Monster leaders started showing up specialized themselves, for example. Monsters had explicit DEX and so did better in surprise and initiative contests. And so forth.
On this, I agree; and I know you're aware I somewhat beefed up the hit points on a lot of monsters to compensate. I also gave them their strength bonuses to hit and damage where appropriate, this is something that always bugged me about RAW 1e monster design that I fixed long before 3e also fixed it.
