"Average PC" is an irrelevant concept. You're trying to pin down the minimum.
That seems a relatively flawed approach to game design. Ignoring the effect an option would have for the vast majority of games in order to compensate for a perceived weakness in the one game where no beneficial feat/item/power/PP/ED choices are made? Not exacty a reasonable approach, in my mind.
Now, tiornys's argument (which you've also briefly mentioned before) is definitely the stronger one - the idea that monster hp has scaled enough that this is what compensates for the enhanced abilities of the PCs.
At the same time, I'm not sure that his example... assuming that PCs are still just sitting there swinging with At-Wills with no other effects in play - is really accurate for an Epic party.
I am dealing with facts, is the problem. Designers designed the system and have public statements about it that flat out disagree with your opinion. Those are the facts of the way the system was designed, so, yeah, it is equivalent to telling an Astronaut the Earth is flat, when he's seen it is round.
No, it really, really isn't.
The only fact at hand is whether a PC loses 3-4 points of attack bonus, relative to monster's scaling defenses, over the course of levels 1 through 30.
That is, yes, a fact.
Whether other elements of the system compensante for that (in terms of PC options) or exarcerbate it (in terms of monster hp) is not something you can directly measure and offer up as objective fact. Similarly, the designers having their own opinion on the matter is not the end-all and be-all; while they may have positions of greater authority on the mechanics of the game, they are not infallible by any means.
The problem with your assumption that character powers help mitigate the 4 pt spread is that some of the benefits you're factoring in only work when the PCs hit. Its kind of a vicious cycle: the PCs are hitting on 14+, but *could* be consistently hitting on a 10 or better if they hit with certain powers. However, they need a 14+ to hit with the buffing powers, which means they don't consistently get the buff. And if some of those powers are Encounters or Dailies, the problem is compounded.
Well, yes and no - sometimes the bonuses do require hitting. Quite often they aren't, if the result of utility powers, Effects, feats, items, etc.
And, especially by Epic levels, many PCs will often be able to bring some resources to bear to ensure a specific attack lands. So if you stack up some temporary attack bonuses for one key attack, which then gives a huge attack bonus to every else for the round... well, it is an effective strategy.
Now, not every group, again, will have that exact approach or those exact powers. But almost every group will have some sort of advantage or approach along those lines, and Expertise only really makes sense when one ignores all those other possible resources.
So why not go with a simple, consistent +1/2/3?
Because it isn't an either/or - you put the +1/2/3 into the system alongside all the other benefits. And so suddenly Epic level PCs are regularly hitting significantly more often than they were at Heroic levels.
Now, the argument that tiornys makes is that they need to be able to hit more often, to compensate for enemy hp, and maybe that is, though I'm not convinced.
And the argument that Aulirophile makes is that it doesn't matter if most groups are now hitting far more often then they should, as long as some idealized 'minimum' group is now hitting at the exact same percentage as at level 1... which I don't agree with either, since, again, taking something like that out of the context of the rest of the game just doesn't make sense to me.