A "doctoral thesis", if you will 
I guess the point is in the classical clash between
possibility and
probability.
Amazingly long range shots did happen in real life. Some pilots claim on kills or inflicitng critical damages from ranges upto 1000 yards and over. Though a certain margin of 'guesstimates' would have to be considered in the fact that none of them exactly had laser-targetted range finders.. perhaps a +/- of 200 yards or so could be accounted for. But that still makes it at least 800 yards or above.
As a technical possibility, the M2 .50s were, as shown in other discussions by the advocates of the reality of long-range gunnery, certainly capable of inflicting damage over very long distances.
However, there also many factors in aerial gunnery which in most cases, very naturally "neuter" the "possibility" so much as to almost consider it "improbable".
According to the renowned Tony Williams, effective range(the range where you can expect to hit, damage, and shoot down a target) of aerial gunnery against fighter-sized targets were at maximum 200meters. Against straight-flying bombers, 400meters.
It is also a very well documented fact that despite the lumbering size and straight flying path, the probability of Luftwaffe interceptors landing a well-placed hit on bomber-sized targets with a heavy cannon was remarkably low. That alone probably shows how difficult aerial gunnery in real life, not to mention that in most cases rarely would people be able to shoot down even a single plane, despite the fact they've encountered and engaged multiple targets in a single sortie.
So my take is that there are two main factors to consider here;
1) still lacking represantation of real life conditions
Despite recent changes in gunnery with AH2, there are still some possible reasons that might reduce the overall difficulty of gunnery in a game, and thus, make it easier for people to hit targets from longer ranges.
For one thing, the distance counter itself gives out very critical information on how much lead is required to hit a target. Despite the various reasons given for the existence of the distance counters, real life pilots seemed to have had as much trouble as game pilots flying with 2D graphical representations and limited resolution/colors.
In other words, despite all the "real-ness", just like game pilots, real pilots had no other way to judge distances than to estimate the range through the crosshairs of their gunsight. The only sure way to place oneself at an absolutley effective gunnery range, was to get as close as possible, since if the enemy plane is right in front of you, at least you know that he is close by.
In sim-game terms, it would be something like this. The limited resolutions and colors of a computer monitor cannot give you any real accurate info on whether the target is gradually closing or departing. However, that being said, there is one way to determine the range between you and the target: and that is, if you would go as close to the target as possible, then you'd know how close you are, since if you get any closer you'd probably collide.
Therefore, IMO, I really don't see why the distance counter (albeit in 200yard increments for hostiles) exists in the first place. I understand the reason for long-range distance indications, such as 6.0k closing down to 3.0k and etc.. however, my opinion is that once within 1000 yards, the distance counter should disappear, and the pilot should only be given the basic icon set, and the +/- sign. The fact that we still can judge the distances pretty accurately, at least by 200 yard increments, is IMO one of the factors which makes long-range gunnery more probable in AH than in real life.
Another factor would be the existence of ammo counters. This is an already well-known phenomenon. When flying planes without any direct means to determine the exact ammo load(as in the IL-2 experience), people usually tend to fire very carefully, at mostly very close distances, since you have no way of knowing how much of ammo you can "spend under control" in hopes of achieving a long-range hit. While this doesn't effect gunnery directly, it effects the pilot pyschological, making him less willing to even try and fire at longer distances than usual.
2) game pilots much more skilled in gunnery than their real-life counterparts
However, another factor which must be considered, is that game pilots encounter, and engage enemies thousands of times more than real pilots. If an avid AH gamer plays the game every day for two hours, meeting 30 enemy planes every session, in a year he meets more than 10 thousand enemy planes. That would be more planes engaged than at least 1,000 real life WW2 pilots.
Although experience doesn't necessarily increase directly in proportion with time, it is probable that an average skilled AH gamer might still be much more experienced in encounters than the an average WW2 pilot.
Thus, it is logical to assume that this would have an effect in gunnery. So it is probably natural to expect people hitting from much longer distances than real life. The problem is, this factor can never be accurately measured, and relies a lot on guesses and casual thinking. It's a largely undetermined variable which may differ a lot from person to person.
...
The trick, is to determine just how much of AH's long-range gunnery is accounted by its lack of real life conditions, and how much is determined by the "pilot skill" factor. While it may be impossible to pin point just how much, there are some empirical figures which might help.
For instance, empirically, the average range of effective gunnery in AH1 was around 500 yards. If an enemy fighter was behind you at 500 yards or so, even if you were in a faster plane it was pretty much impossible to expect you to out-accelerate the enemy plane and pull away to safe distances. 600 yards+ was considered tough but possible, and 800 yards shots were considered rare but still too often to consider "improbable".
When simular series of discussions concerning reallistic gunnery arose in AH1 days, the same arguments relying on the "pilot skill" factor were given.
However, the advent of AH2 and its more refined "hit detection resolution" quickly proved that might not have been the case. Everyone more or less still remembers the initial "shock" of AH2 beta. People suddenly finding out that they can't hit as well as used to, was more or less a proof that in AH1, people were getting hits that should have been a "miss". It became much clear that a large part of the long-range gunnery in AH1, was due to an out-dated hit detection system, and not really as much a "skill factor" as some people would have preferred to believe. I would say that it was a small, but very significant turn of events which proved that people have a way of justifying the current trends to be "normal", when in fact, it was caused by a faulty(or rather, out-dated) system.
After the new gunnery has been established, I think it would be safe to say that in AH2 a plane that might be able to outaccelerate the enemy behind it, has a fair chance to pull away to safety for a duration of anywhere between 5 seconds to 10, 20, even 30 seconds, despite getting a few hits, even at 400 yards(indicated) range. Were it AH1, in such a situation, the targetted plane would be shot down within 5 seconds.
Although long range gunnery still happens, especially with those with armament much more suited for it, still AH2 gunnery has corrected a lot of the problems and brought the gunnery much closer to the real thing than before.
If there is anything to add, I would be highly interested in how things turn out if the distance counters vanished under 1000 yards as proposed, and the ammo counters were deleted from planes that did not have them. When some people still get 600~800 yard shots despite those changes, then I'd openly praise
THAT, as undeniable work of skilled gunnery. However, I have my doubts.
