It's not to say that they aren't valid, but they are a critique of the kind of game that OD&D is, not of broken rules. D&D is not heroic fantasy, it's low fantasy, and it's not a game of power fantasies for each player, but a a game of collective world-building between players and GM. By "power fantasy" I'm not trying to be dismissive to other games or genres that are about individual advancement of an avatar, I'm attempting to draw a distinction between a fantasy narrative that is of individual success (empowerment) and one that is the narrative of a world (like a history).
Old style D&D is not a story about any one PC, it's a story about the adventuring party as a whole, or at another level a fictional world as a whole. That's why replacing characters is so easy. That's why power levels are relatively compressed and monsters randomly encountered and not tiered. At an even more abstract level it's about the game world and slowly revealing it in cooperation with the GM.
|Model OD&D Adventuring Party|
1) D&D is first a cooperative game.
The players will only survive if they cooperate, and the better they plan and cooperate the better they will survive. Yet, it's also cooperation between the GM and the players. An interactive story told by the GM, interpreted by the players and then changed by the GM to suit those player's interests. This is why OSR gamers tend to fetishize player choice and complain about railroads and quantum ogres - both of these things detract from the player's ability to help weave the world.
The rules subtly encourage Player world building via the XP system. Characters all seek to grow in survivability, and may only do so by plundering and then spending cash. They move wealth from ruins/lost places into a game economy. Since the characters can only advance by spending large amounts of gold they do large things with it.
In the ODD game I play on G+ PCs have: tried to invent spells, built hidden shrines to dubious gods, bought property, erected memorial statutes and now at 6th level are attempting to rebuild a road with a fortified toll post to skim off new trade. None of this was planned by the GM, the world was just open for it, and even if our party gets wiped out by the forces of barbarism that object to new trade routes - we'll leave something behind (in addition to a treasure trove of equipment). If the plan succeeds it'll be a trade route and tower and if it fails a 1/2 built road, an army of dead mercenaries and at least one impoverished town. It is also why a Total Party Kill is a good thing. The destruction of the characters becomes part of the world.
Example of a world building TPK: When the random encounter dice show that the emboldened 2nd level party is ambushed by thirty longbow wielding bandits, and the party is surprised, loses initiative and falls to a swarm of arrows without a chance to respond, the game world gains something. Any GM that fails to then mark a bandit hideout with an especially effective and cruel gang of marauders in that hex is making a big mistake. Future parties will avoid that road, unless they are seeking out the bandit band. If the bandits are ever killed they should obviously have some of the dead party's stuff.
1) Your PC is bland for a reason.
A first level PC is a line of stats and has a fair chance of dying on his or her first adventure. This is because OD&D is not about great heroes with destinies. The adventurer is a bumbler: a down on his luck ex-soldier, a disgraced wizard's apprentice, an exceptionally greedy thief or an overreaching religious fanatic. Seeing the party for the first time and not thinking "Some of those mooks are not coming back" is the right response. Your character can become a hero, a mysterious individual with a back-story and power over time and through play - but you start out as another guy in a battered suit of chainmail with a sword. One line or even a few words of description is good enough to build on.
Check out the excellent character generator at Save vs. Total Party Kill - everyone should be using this because it creates interesting, vaguely defined and ultimately disposable PCs (though it has a tendency to create creepy children). Sure you may not have wanted to play a mature, obese elf in a threadbare robe with the spell hold portal, but seeing how this character will interact with the world is part of the story the game is creating. Over time 'Rain Eyes the Rotund' will develop. If she somehow survives at 10th level the threadbare robe will be replaced by something strange, the game world will undoubtedly provide a personality through play, and you're pretty likely to have an interesting character. If not a hero, then at least someone with a compelling back-story and a few unique traits (flame trap scars, a golden animated hand and an imp companion are all likely). What the character also has in a whole mess of dead companions. This is a good thing, from Groog the fighter - eaten by the first rat ever encountered, to the elf's much missed magic-user buddy "The Shrouded Necromancer (Formerly Chumly)", whose spell books she now possesses. These losses also form a compelling character arc, because they aren't just a background element about the death of unnamed friends, and they create a background of fictional deeds the player will remember and act based upon with real knowledge and engagement.
This becomes extra important when the GM is smart and lets evolving player eccentricity create world fiction. The character that decides to follow a minor god and makes up a fiction about that deity gives the GM's world a new cult (and how much the character's beliefs correspond with the GM's version of the god/religion is up to the GM - and provides quest hooks or makes the PC a lone crazy), every band of goblins spared and taken into service creates a new sub-race for potential replacement PCs and every new spell researched can be stolen by or sold to another wizard.
The flip side of this is that simple starting characters discourage drama. Players that want to play a disruptive type still can, but if the PC survives it will only be by working with the party, and they have no elaborate back-story to justify acting like a jerk. If they develop back-story in game (insanity or evil say) of the sort that can derail play it's as part of the adventure and the other players will likely have figured out how to adapt or have some sympathy.
3) Even as characters grow in uniqueness, it's the party that's important
As a companion to the above thought, it's not really the individual character at all who is the narrative agent, or focus, of the game. The adventuring party is the focus. Even if the member characters change, it's the same party and it's the collective efforts of the players that writes the world fiction, only rarely the ideas of one player or the GM. Characters in OD&D don't split up and run all about the world solo, they make collective decisions to travel, delve certain dungeons or battle certain factions.
If one looks at the game from this prospective, the problem of arbitrary death is not a problem, it's integral to the game. Deaths insure that the party will evolve and change over time and tell a story of its own. Without these character deaths the danger of the world is never realized, and narrative will never have the tension of defeat. It's key to understand that as a cooperative exercise in fantastical world building, individual residents of that world are most often not necessary to that larger story. In a sense OD&D's punitive character death rules help make sure the game is not a set of often conflicting individual fantasies of the players about their individual avatars. The ability to sit down with a few other people and spin a fantasy world is what makes table top games uniquely fun, and I think that OD&D's system for doing so emphasizes this aspect rather well.