I'm going to be the voice of dissent here... There's always one, eh?
It seems strange to me that the same people who say that a tighter bore improves accuracy, can also say that the length of the barrel either doesn't matter at all or doesn't matter much. It is friction between the inner bore and BB which imparts additional directionality to the trajectory of the BB or, in pure mechanical engineering terms, the barrel exerts a force on the BB which alters the sum of the BB's angular momentum, relative to the rest position (the hop up chamber) and the centre of the earth (bearing in mind that no matter how level a shot may appear, it is actually on a ballistic trajectory which is some shape of arc relative to the imaginary straight line that would be the shortest distance between the rest position and the target) and this is then further altered after the BB exits the muzzle by air resistance, wind pressure, gravity and back spin to produce a constantly changing vector (direction, relative to an imaginary still point, and velocity).
Thus we see (in the somewhat pompous language of engineering equations) that if the additional frictional force exerted upon a BB by a tighter bore (which incidentally is compensated for in terms of simple speed by the additional pressure developed behind the BB due to the smaller gap between BB and barrel, but the direction of the force nonetheless adds to the sum of angular momentum) results in greater accuracy, then it follows that a longer barrel, which will also generate more friction between the BB and inner bore than a shorter, will also result in greater accuracy, even though it will take additional power and/or a better air seal at the cylinder end to produce the same velocity as from the same length tight bore barrel.
To be strictly accurate, I should mention that the frictional force applied by the barrels actually has multiple directions, each of which are summed into the BB's angular momentum. In fact it is these very multiples that cause the BB to fly "straighter": a little bit left, a little up, down, left some more, right, down, right, up, etc. sums into a vector, at the point of exit from the muzzle, that is closer to an extension of a line from the rest point through the foresight, on a flat 2D plane, than it would be with fewer multiples...
All of which probably explains why, in my own experience, an ICS M4 is more accurate than a G36C (Ed, what make are they @Skirmish?), which in turn is more accurate than an AK47 Beta Spetznaz, but my CYMA AK47, with about 95mm more barrel, is more accurate than the ICS M4, and my Ares L1A1 SLR with an approx extra 165mm of barrel, is more accurate than my AK... however I'm not sure that Daft Lad's L96 with an extra 30mm more barrel is even as accurate as my SLR, despite having better optics that are also set up properly, whereas so far I haven't sussed how to get my 20x32 scope working well for me (not that I've actually put much effort into it yet though :lol: ).