Li Shenron
Legend
Whether I agree or not, these are good points to bring up! 
This is very group-dependent for me. In our tabletop games, no one has ever multiclassed in more than 2 base classes and 1 prestige class. That sounds still very decent to me.
OTOH in oneline games, there's lots of players who focus more on character build than the actual adventures. It is sometimes irritating to see combat-types with 5-6 classes.
Maybe that's just me, but I don't buy the idea that "character concept" requires heavy multiclassing. I think that every concept can be covered by 1-2 classes at most if you have a large number enough of feats & spells, or eventually if you ask your DM for a VARIANT of the class in question. For some reason I cannot explain, people prefer to play with multiclassing (and often complain that they STILL can't get the perfect thing they wanted) rather than going straight for the target by creating a variant of the class with the DM.
Anyway, I agree with you that the XP penalty should have been designed better. It could have been SMALLER but EASIER (like e.g. 5% penalty for every class after the first, doesn't matter what level).
Agree completely. Metamagic is great for spontaneous casters, period.
There is no reason why metamagic MUST be good for every caster: spontaneous casters are all about flexibility, preparation-based casters are all about knowledge.
Of course, having an OPTION is good, but 3ed took IMHO a bad path when they decided to make metamagic (very nearly almost) just as good for preparation-based casters. And at the same time they DIDN'T do the same service to spontaneous casters with options to improve their spell knowledge (Extra Spell feat can not be seriously compared in usefulness).
It's not a matter of balance really, it's a matter of taking away someone's unique ROLE, and also not giving much in return. No wonder if there are nowadays probably more Warlocks than Sorcerers.
Agree, but the best solution here is the DM.
Agree partly. Fortunately I largely play 3.0 (so at least PA is much better) and I also always enforce my own control of magic item's availability for sale. I still think that even in 3.0 TWF is more costly than it's worth, but otherwise 2Handed and Sword&Board warriors are quite on par.
Well it's hard to say. Sometimes I want adventures where there is 1 encounter per WEEK, and sometimes I want a single scenario with lots of battles in a single day.
However I definitely enjoy how the game has multiple levels of strategy and resource management, hence I dislike almost every variant rule or option that eliminates one of those levels.
I don't know... if all buffs are short, people are still going to cast them shortly before combat, and it may actually become more boring than having a few of them cast in the morning and permanently added to your character sheet.
Clearly, long term buffs should be of less power than the short term ones.
I never had a problem with this (tho the 3.0 PA is way simpler than the 3.5 version). Usually we just choose a number without thinking too much about the exact probabilities...
I like both ways. In a game with no munchkins I prefer rolling and accepting the results (just for fun), but otherwise point buy is safer from party inbalance.
Just kindly ask the players not to take multiple abilities of this kind.
For PC I think it works quite well. My possible only gripe is that you inevitably end up thinking in terms "I have to out and kill some orcs, so that I can finish crafting this item" :\

airwalkrr said:1. Multiclassing
AD&D was too strict. 3rd edition is too permissive. In AD&D you couldn't change professions unless you were human; that was a problem. In 3rd edition you can't help but change professions many times, including certain prestigious professions; this is also a problem.
Potential Solution: There needs to be a significant drawback to multiclassing while keeping it viable. Currently, the only drawback (XP penalty) is easily circumvented. Even when it isn't avoided, XP loss isn't fun.
This is very group-dependent for me. In our tabletop games, no one has ever multiclassed in more than 2 base classes and 1 prestige class. That sounds still very decent to me.
OTOH in oneline games, there's lots of players who focus more on character build than the actual adventures. It is sometimes irritating to see combat-types with 5-6 classes.
Maybe that's just me, but I don't buy the idea that "character concept" requires heavy multiclassing. I think that every concept can be covered by 1-2 classes at most if you have a large number enough of feats & spells, or eventually if you ask your DM for a VARIANT of the class in question. For some reason I cannot explain, people prefer to play with multiclassing (and often complain that they STILL can't get the perfect thing they wanted) rather than going straight for the target by creating a variant of the class with the DM.
Anyway, I agree with you that the XP penalty should have been designed better. It could have been SMALLER but EASIER (like e.g. 5% penalty for every class after the first, doesn't matter what level).
airwalkrr said:2. Free Metamagic
Metamagic as an idea is great. But there is a reason the designers gave it an opportunity cost in the form of a higher level spell slot. Wizards were not meant to maximize fireballs at 5th level and clerics weren't meant to chain greater magic weapon at 7th level. The current trend of metamagic rods, sudden metamagic feats, and other "free" metamagic effects gives spellcasters too much power.
Potential Solution: I think this aspect just needs to be removed from the game. Metamagic is fine. "Free" metamagic is not.
Agree completely. Metamagic is great for spontaneous casters, period.
There is no reason why metamagic MUST be good for every caster: spontaneous casters are all about flexibility, preparation-based casters are all about knowledge.
Of course, having an OPTION is good, but 3ed took IMHO a bad path when they decided to make metamagic (very nearly almost) just as good for preparation-based casters. And at the same time they DIDN'T do the same service to spontaneous casters with options to improve their spell knowledge (Extra Spell feat can not be seriously compared in usefulness).
It's not a matter of balance really, it's a matter of taking away someone's unique ROLE, and also not giving much in return. No wonder if there are nowadays probably more Warlocks than Sorcerers.
airwalkrr said:3. Synergy
I'm not talking about synergy bonuses from skills. I am talking about unintended consequences of mixing and matching sourcebooks. For example, if a dread necromancer (Heroes of Horror) takes the Tomb-Tainted Soul feat (Libris Mortis), he gets unlimited healing. Taken separately, neither of these abilities is overpowered. Taken together, they have synergy that is far greater than the power of their individual components, likely an oversight because the books had different development teams.
Potential Solution: Allow each player access to one sourcebook ONLY outside of the three core rulebooks. This prevents most forms of synergy. Of course it reduces player options so it is not an ideal solution.
Agree, but the best solution here is the DM.
airwalkrr said:4. Two-Handed Weapons
Because of the double bonus from Power Attack, floaty shields, and more beneficial Strength modifier, two-handed weapon wielders have become the staple of melee combat. Forgive me, but this is trite. Two-weapon fighters and sword-and-board style have become comparatively worthless relics in the game.
Potential Solution: (and this is vague) The game needs to support multiple fighting styles by providing viable options for each that don't heavily overshadow the others. Player's Handbook 2 goes a long way towards rectifying this problem, but I don't know if it goes far enough.
Agree partly. Fortunately I largely play 3.0 (so at least PA is much better) and I also always enforce my own control of magic item's availability for sale. I still think that even in 3.0 TWF is more costly than it's worth, but otherwise 2Handed and Sword&Board warriors are quite on par.
airwalkrr said:5. Balancing Per Encounter Instead of Per Day
This is a horrible idea because it propagates the notion that the world conforms itself to the power level of the player characters. Some encounters are meant to be tougher, and those encounters require greater resources. Others are meant to be more menial and require fewer resources. Properly gauging the difficulty of an encounter and balancing your resources is part of the strategy of D&D. Leave "per encounter" balancing in MMORPGs and keep D&D a strategic game, like it was meant to be. Or at least publish two versions.
Potential Solution: (another vague one) Characters should not be able to use their most powerful abilities without limit.
Well it's hard to say. Sometimes I want adventures where there is 1 encounter per WEEK, and sometimes I want a single scenario with lots of battles in a single day.
However I definitely enjoy how the game has multiple levels of strategy and resource management, hence I dislike almost every variant rule or option that eliminates one of those levels.
airwalkrr said:6. Neverending Buffs
Yet another thing that removes an element of strategy from the game. Clerics are particularly fond of these. Spells like magic vestment, greater magic weapon, and heroes' feast are virtual must-haves for clerics because they last practically all day, especially with extend spell. "Forget situational spells. Just memorize the ones that keep you perpetually powered-up!" That's bland.
Potential Solution: Reduce durations of spells like this, or add costly material or XP costs to reduce their frequency of use.
I don't know... if all buffs are short, people are still going to cast them shortly before combat, and it may actually become more boring than having a few of them cast in the morning and permanently added to your character sheet.
Clearly, long term buffs should be of less power than the short term ones.
airwalkrr said:7. Combat Expertise and Power Attack
These kinds of feats make the game a bit too complicated because of the constant calculation required. A 10th-level fighter with Power Attack has 11 attack options representing the various penalties he can take. A 10th-level fighter with Power Attack AND Combat Expertise has 66 attack options! And he is expected to quickly decide which course of action is best?
Potential Solution: Simplify these kinds of feats with a flat penalty and flat bonus. The decision for the player then becomes merely to use it or not.
I never had a problem with this (tho the 3.0 PA is way simpler than the 3.5 version). Usually we just choose a number without thinking too much about the exact probabilities...
airwalkrr said:8. Point Buy
As if we needed more excuses for players to focus on character creation as opposed to actually playing the game. The world isn't that fair. I don't know why we would expect our characters to be "equal" either (as if that ideal were even possible). Besides, it ruins the excitement of rolling up a really nice set of scores.
Potential Solution: Roll ability scores.
I like both ways. In a game with no munchkins I prefer rolling and accepting the results (just for fun), but otherwise point buy is safer from party inbalance.
airwalkrr said:9. Rerolls
Various class abilities that allow rerolls greatly reduce the amount of chance in the game. You aren't likely to roll very many 1s during a game session, and if you have one or two reroll abilities (luck domain, luck blade, fatespinner, etc.) you don't need to worry about them. As long as your character is powerfully built, you will almost never need to worry about pesky automatic failures. Additionally, these abilities are greater still in the hands of NPCs, who only usually need them for one battle.
Potential Solution: Don't allow rerolls to change the result of automatic successes or failures.
Just kindly ask the players not to take multiple abilities of this kind.
airwalkrr said:10. Magic Item Creation
It costs XP to make magic items. So my character unlearns things for succeeding at a task. How on earth does that make sense?
Potential Solution: Just drop the XP cost for magic item creation. It already costs your character a feat. Or make the creation of magic items difficult by requiring rare components that must be quested for.
For PC I think it works quite well. My possible only gripe is that you inevitably end up thinking in terms "I have to out and kill some orcs, so that I can finish crafting this item" :\