It is fine if the players are all about as good at gaming. In my old group, there were two guys that always won shit like that, so it got old a long time ago. I try to only run team games for them.

There's a point I hadn't considered too deeply, that some players are just better at making their characters, and would therefore win more often. I'm up for a slight imbalance, but if it's too overbearing, that would cause friction within the group.

I've fenced against two-weapon styles a few times (dual dagger, dagger and rapier, cloak and rapier), so I can vouch for it usually being about one for defense and the other for offense, but...

Using two weapons, like using one weapon, is a game of cunning and tactics. Unlike a shield, having two weapons means if you defend to the right you're free to attack to the left and vice versa. It's sort of like hitting high then low, but you can hold your opponent's weapon at bay.

Of course, this tactical element of combat is lost in every rpg I've ever played, because it's abstracted into the die rolls. If your mechanics work in a way similar to most traditional rpgs, I'd suggest that beginner two-handed fighting gets a defense bonus, medium skill two-handed fighting a to-hit bonus (parry to create and opening and then strike) and only experienced fighters would get an extra attack.

I'd also note that off-hand penalties aren't as problematic as you might think, that a little practice minimizes the difference. I'm far from ambidextrous, but opponents have trouble figuring out which hand I prefer when fencing.

Terribly sorry about that. My pet peeve is actually misplaced nostalgia. I work in an antiquarian bookshop and deal with misplaced nostalgia all day.

Anyways, I'll try to stay on topic. I see games that play out quickly and games that are neat as being two different problems that sometimes overlap.

Games that play out quickly are not just a matter of simplicity vs. complexity. Some simple games run out of possibilities (tic tac toe, Pantheon from Hogshead), but a simple game, well constructed, can produce a lot of possibilities (Go, or Over the Edge). A poorly designed complex game can sometimes have only a few combinations that actually work (MMORPGs), or sometimes have a lot of options that do the same thing (the HERO System is surprisingly bad for this). On the good side, I've found that Legend of the Five Rings produces a tremendous amount of play for those who wade through its complexity.

Yeah, the predictability of the WoD pattern is innappropriate to the complexities of the subcultures it presents, but I think the pattern is really there to create a play dynamic. It helps ensure at least 2 people in each faction for a 12-15 player turnout at a larp. The question is, are you willing to accept some artificiality for a play dynamic?

So, how to create long-term interest? I think that adaptability is important. Can the game handle something unexpected, can it adapt? There will always be something that falls outside the game's scope, but the game should have patterns which suggest how to create and integrate whatever you need, both in terms of setting and mechanics.

D&D is the most popular game out there and even its detractors have nary a peep regarding the races aside from 'too tolkeenesque', which is hardly and objective measure of quality.
...
Each 'fatsplat' has five 'races' and five 'factions'. Fatigue sets in when faced with a new 'fatsplat'.  The Races (in vampire this is most explicit) do divide up much as I postulate above. Gangrel are tough (Resiliance as a clan discipline), the Nosferatu are strong (vigor) and so on. There are actually nine attributes and thus there should be nine clans, but that isn't quite the pattern being filled.... but bear with me.

Compare the old world of darkness with its original idea of 13 clans, which was then complicated by the spread of new, additional clans scattered here and there, and the fact that only... what? 9 clans actually appeared in the main book?  Then there was the crossover on disciplines. If you liked a particular discipline (celerity, for example) you probably had two, if not more, clans to look at in the main book alone (Brujah and Toreador in this example).

Actually, there were 7 clans plus clanless in the original main book, representing the 7 deadly sins and the search for a new deadly sin. But yes, it turned into a mess.

Yet, even after five years it is quite easy to find players who have not switched over to the new rules/new setting.

I think that's because of something separate from the cleanup of the mess (a mess which was rapidly reintroduced within months of the new edition). The new edition is not evocative. Whether you liked metaplot or not, just the idea of an impending apocalypse added more to the game's morality play than any of the archetypes. The nWoD world is just...blank. The morality play that defined oWoD has been pared back to the bone. The elements of seduction (the best selling component of current-day vampire stories) has also been downplayed a little, reducing the evocativeness further. What's left is politics, which can be fun, but it's less than what oWoD offered.

Comparing to my own older thoughts on the subject it occurs to me that while an elegantly designed rule set and world  can be very appealing aesthetically it may actually be poor design.   Too elegant a rule set may be too easily mastered, leading to boredom with nothing new to find out, it certainly leads to 'game exhaustion' when talking about a 'house engine' used for multiple games in a line.

Actually, calling nWoD elegant is a bit like calling a battleship maneuverable. However, nWoD offers more options than oWoD thanks to emergence, which can be an elegant concept. Since there are few constraints in either nWoD or DP9, I have to assume you're referring to feeling constrained to emergent archetypes. A game may allow you to play anything, but only allow a few templates (emergent archetypes) to be successful (MMORPGs are horrible for this). Yep, it's a problem. One way to solve it is to not over-focus the game on one task (e.g. combat, or politics, as in nWoD) and include other activities (seduction, moral debate, religious debate, as oWoD had).

Yet, despite a near burning rage at level/class based design, I can still be convinced to pick up a new D&D based game just to find the new, messy, ways they do things.

And you're not alone. D&D is made accessible due to its familiar terms. Fighter, elf, dwarf, these are words that even non-gamers often recognize. Lancea Sanctum, Daeva, these present a steep learning curve for players to get into nWoD. That means D&D's archetypes are easily grasped, while nWoD players must learn everything from the ground up before starting to make a character. It discourages exploratory play, and that's a lot of rpg players.

Just a random train of thought, but it does occur to me that in the long run an excessive elegance is probably just as bad for a game as an excessive mess... which in itself isn't relavatory but: the line for 'excessive' in elegance may be much lower than recognized.

If you want a game that has excessive elegance, look up Go. It's had a short run of...well, it's estimated in the thousands of years. The problems with nWoD come not from elegance (or emergence) in archetype construction, but from the exclusion of many side thematic elements in the game. The game's personality is muted.

Simply put, nWoD is boring. oWoD is not, so people put up with the mess.

I definitely have more fun now than I did back then. The challenges I face or set up in games are a lot more interesting. I even get that sense of wonder sometimes, but that's because I don't play the same rpgs anymore. The gamebooks of the rpgs I played twenty-odd years ago might as well be firewood, even the new editions of those games.

If I want those old feelings back, I need new games to challenge and explore.

I don't think I made it through a year (though it's a little hard to remember that far back now) before I tried a second game. The pattern since trying that second game, which I think was Gamma World, has always been to try new games. At last count I think I was over 60 different rpgs that I could remember, and that's with certain games grouped (e.g. all World of Darkness variants counted together as only 1 rpg played).

My frustration with D&D4e is how it's so solidly adhered to its story formula and its attempts to hide this under the presumption that "roleplaying" around the rules fills any gaps. What's worse, there is within that the suggestion that anything else is of little or no consequence, since the mechanics do not give them weight (strategy, investment, tension, reward, etc.). If the premise of the game was also limited, then I could accept that, but it isn't. It suggests a complete world, but the rules cannot cover 99% of that world or the activities (and story possibilities) in it. You're supposedly an adventurer, but adventuring is about trying new things. D&D4e is about repeating the process of killing and looting over and over. That's really limiting, both as a player and GM. For this reason, I find D&D4e to be the most pretentious game I've come across in a long time.

Frankly, I like hit points, but it has little to do with either side of the argument presented. I like how hit points (and healing) tie together the length of adventure and the rough magnitude of the characters' fame (or infamy).

As characters increase in levels, there is a balancing act between damage taken and the pace healing (including healing spells) can be used to recover. Individual battles may not be threatening, but if the damage accumulates faster than the healing can cover, tension increases. Eventually, the climactic battle pushes the characters as close to their limit as the DM dares.

What this means is that minor heroes (low-level characters) have short adventures, reflecting their limited reputations. They couldn't survive longer, but they also get quicker rewards, too. When characters become the heroes of the realm (high-level characters) they also have the hit points and healing to undertake long quests befitting their status. There is more invested in the characters by that time, allowing longer times and more work between rewards.

This could be done with more complex damage-tracking systems, but it is so much easier to do with hit points.