The 57mm was far more accurate than RPs, but was less flexible in that RPs could be fitted to just about anything and didn't need a special model of plane. This is from 'Flying Guns: World War 2' by Emmanuel Gustin and myself:
"Finally, it is worth mentioning the other big gun used operationally by the RAF, the Molins 6pdr fitted to the Mosquito Mark XVIII, or "Tsetse". This was designed as an automatic anti-tank gun for the army but subsequently developed for aircraft use, still as an anti-tank weapon. The massive recoil was absorbed by an 80 cm recoil movement, which as with the Class S kept the peak loads very low. The RAF ordered 36, the first being delivered in July 1943, with a total of 27 being constructed. Before they saw action, requirements changed yet again and they were handed over to Coastal Command for anti-submarine and anti-shipping duties, so they were never used in the anti-tank role for which they had been designed. They undoubtedly would have been effective in terms of both power and accuracy. The 2.8 kg shot, fired at 890 m/s, was capable of penetrating 82 mm / 450 m / 60º and 71 mm at 900 m when fired on the ground; obviously it would improve on this when fired from the air. A War Office report stated that four rounds were normally fired per attack, of which an average of 1.3 could be expected to hit a 3 m2 target. This is equivalent to a 33% hit rate against a tank, usefully higher than the Hurricane IID (presumably because of the fuselage mounting and much higher muzzle velocity)."
By comparison, RPs had a 5% hit rate against tanks in training, and 0.5% in combat. The reason why the Tsetse wasn't used for ground attack was probably that it was considered too vulnerable to ground fire because of its size.
Tony Williams: Military gun and ammunition
website and Discussion
forum
