Who is saying that PCs can't have weaknesses? If you've got 4e in mind, PCs have weaknesses. The fighter is weak at ranged combat and socially. But probably fit. The wizard is weak in melee and mediocre socially. But probably learned. The ranger is strong at either melee or ranged, but typically not both, and is weak socially. But probably strong in exploration.Why can't my character have waeknesses and strengths? Why can't I make a non-combat character without breaking the system?
If by "weaknesses" you mean "inability to effectively/meaningfully contribute", that's a different matter. D&D has some distinctive features - its very strong emphasis on party play, and its typical stakes being "win or die" - that I think make an inability to meaningfully contribute tricky for any major field of activity. And because D&D has always, mechanically, prioritised combat as a site of conflict resolution, non-combat PCs will tend to "break the system". Of course, if you play the game ignoring the bulk of the action resolution mechanics, then a non-combat PC might be viable (and 4e offers a novel, metagaming take on the non-combat PC via the "princess warlord" build).
But personally, if I wanted to run a game which wasn't focused on combat as a primary site of combat resolution, I wouldn't run D&D. I'd look for a game which doesn't place so much emphasis on combat in PC build and action resolution.
Mechanical effectiveness can extend beyond combat, though. Assuming there are non-combat action resolution mechanics.To a certain degree, "mechanical" balance in terms of combat effectiveness is only really a priority in design if you believe that your players treat combat effectiveness as a primary vehicle of "fun" play, or place a high priority on the emotional satisfaction they receive by being combat effective.
I agree with this. "Balancing across pillars" requires, at a minimum, that each of the pillars be meaningfully present in the default mode of play. To date, what they have said about this (in L&L, in the playtest documents) has been a bit disappointing.If D&D Next really manages to bring the 3 pillars to shine (combat, social and exploration) and be equally relevant, then I think it's perfectly fine if the Fighter dominates combats. Such a system will probably mean that even tougher combats are over in 15-30 minutes at worst, and that's probably a "spotlight" time that every group could deal with. And it would imply balance, if a 15 minute fight is followed by a 15 minute exploration where the Rogue shines and a 15 minute social interaction where the Bard shines or whatver.*
But if it works like 3E, Pathfinder and 4E, and combats can take one or more hours, and many adventures containing multiple combat encounters followed by an exploration or social "piece" as binding glue, then a Fighter shining in combat and everyone playing second fiddle will not be much fun.
Given that 4e is expressly designed so that PCs have quite different damage outputs, I don't think this is true: the fighter in my 4e game does 1d8+13 on an at-will hit; the sorcerer's at-will burst 2 does 1d4+27, or nearly twice the damage against mltiple targets. The difference is that the fighter exercises a huge amount of battlefield control, whereas the sorcerer has comparatively little.Better at what?
Fighting? Diplomacy? Stealth? Ancient Lore? Spell casting? Riding? Climbing?
From what I can tell, your definition (and many if not most 4e-enthusiasts' definition) of balance is, "Total balance between characters in combat damage output and battlefield effectiveness."
More generally, who upthread from the "4e/balance" camp has defined balance in terms of combat effectiveness? People talk about mechanical effectiveness; why assume that that is limited to combat? The invoker in my 4e game has two skill training feats, Linguist and Arcane Familiar to get a Book Imp: none of those are combat abilities. They're about making the PC the preeminent scholar in the party, and probably in the gameworld also.
If by "unbalanced" you mean "breakable by a horde of min/maxers", you are describing virtually any game.
I don't think HeroWars/Quest is vulnerable to abuse, or being broken by hordes. RuneQuest and Traveller are both fairly resilient, also.If anything, an rpg system that can't be abused isn't much of an rpg.
Part of the issue is how far you have to drift the game to break it. One of the issues with 3E is that some players break the game without trying, just because they try out the stuff that a druid, wizard or cleric can do, under the (as it turns out, mistaken) apprehension that they're playing the class as written. Heck, in the second session of 3E that I ever GMed we discovered that Summon Monster was broken - with a 3rd level spell the PC could either summon a Thoqqua, or summon a Triton that could then use its own Summon Nature's Ally to summon a Thoqqua. And that wasn't even trying - that was just a player looking at his PC's spell list and the corresponding monster stats.
At a minimum, if the game is going to break that easily I want (i) the designers to tell me where the break points are, and (ii) to show me where they think the viable game can be played without hitting those points. Burning Wheel does a pretty good job of this, so it's not like WotC have to come up with the idea all on their own.