(Sorry I didn't elucidate earlier; I had guests show up and games to play.)
We're talking, as far as I could tell, about whether or not the system can tolerate a large initial spread in party AC. To that point, it really becomes incumbent on those players whose PCs have low AC to take some large part of the responsibility for mitigating that onto themselves.
I beg to differ.
Let's take the AC 13 Dex 16 Sorcerer. This player went out of their way to put 5 or 9 starting points (race dependent out of 22) into AC. The Paladin put zero starting points into AC, but has an AC of 20. That's one player that put 22% or 41%, a hefty chunk, of starting points into it and the other that put 0%.
There's a flaw here.
To now say "Hey Mr. Sorcerer, why don't you take Leather Armor?". So, the Sorcerer takes Leather Armor and has an AC of 15.
The Sorcerer still has a lower AC than nearly every single other pre-feat class with the exception of a Shaman, even though the Sorcerer took a feat.
On top of this, the first levelSorcerer doesn't have a lot of defensive powers since the Sorcerer is a Striker. At first level, the Sorcerer has one Daily power that might be considered a bit defensive as well as offensive if it dazes a foe or some such. And, this Sorcerer is also disincentivized to take powers like Bedeviling Burst, Tearing Claws, Tempest Breath, and Thundering Roar because it brings the PC too close to his foes.
That's it. The Sorcerer went out of his way to be a bit defensive and is still way behind the curve. The Paladin (Fighter, etc.) did nothing and are way above the curve. Don't get me wrong. The Defenders should have high ACs, I'm more concerned with the fact that there are classes that go out of their way to up their ACs and may or may not make it to an average AC.
I'm concerned that there is a 7 difference when a player goes out of their way to boost Dex between best and worst without a feat.
If the player puts in Dex 14, than the difference is 8. Then, same level foes hit that PC on a 6 and N+4 foes hit that PC on a 2. Say what? A suggested potential encounter and the PC gets hit on a 2? By game design? With Dex 16 and Leather Armor, the PC gets hit on a 5 in that case.
So, this is really all a long-winded and rambling way to say that a large relative low-high AC spread isn't intrinsically game-breaking; it just needs more carefully considered tactical play. And the tools for this play can reside (1) within the builds for the entire party and (2) even within the builds for the single characters with low AC.
I don't see it. Combat is chaotic. The best laid plans of the PCs (and the DM) go askew all of the time.
In fact, one of the frequent things I hear in my home game from one player to another or to me is: "Oh no, you ruined my plan". My players are always thinking about what to do next and a monster moves, or a PC slides a foe, or a PC moves into a group of foes, or whatever.
Combat is chaotic. The chances of an AC 12 or 13 or even 15 Sorcerer not getting into the brunt of an encounter at some point is virtually nil. At least if the DM runs the quasi-intelligent monsters as quasi-intelligent.
Your point is that AC is not the entire picture. I agree. But, the static defenses like AC and NADs are a very large part of the picture since they are not dynamic. The dynamic elements change. For example, concealment for a Warlock. There will be times when that Warlock is dazed or stunned or immobilized and he cannot move. His normal defenses just dropped by 2 in addition to any combat advantage.
So, the static defenses are very important. The dynamic elements of the game system do affect this, but to a much lesser extent.