From a translated 1943 Italian document (originally written by the Germans?);
Development of Sights for Fixed and Free Guns which detailed German efforts in this area up until that point.
Dispersion due to the aircraft itself is much greater. This has been experimentally determined by installing in a fighter aircraft a camera gun (all errors if a ballistic nature are thus cancelled out) and choosing the simplest form of attack, i.e. prolonged burst, exactly on the tail of a bomber aircraft on a straight course. In this form of attack the operations of aiming are very simple, as it is sufficient to sight slowly and fire only when it is certain that the target is centred in the sight.
The leading aces of the German Fighter Arm were chosen to carry out the experiments and camera gun tests were carried out for ranges of 500 m. and 1000 m.
The results were discouraging. The winner of the competition, Oberst Galland, got the best score with an 8% dispersion i.e. only 50% of the rounds at ranges of 500 and 1000 m. lay within circles of diameter 8 m. and 16 m. respectively.
Such inaccuracy is inevitable in air gunnery, since it is bound up with the instability and uncontrollability of the platform represented by our aircraft, which in the tests under consideration were the Me.109 and Me.110 i.e. two of the most manoeuvrable types of aircraft.
For what it's worth British testing at the Gunnery Research Unit found that on average a pilot could hold his sight within 0.23 degrees (3ft at 250 yards) 50% of the time.
Compare that with 0.92 degrees of Galland -- a pilot certainly not representative of their average ...