The shell itself flies considerably longer, although 20mm shells start to slow down and drop much more quickly than even slightly bigger shells like 30mm (see e

The shell itself flies considerably longer, although 20mm shells start to slow down and drop much more quickly than even slightly bigger shells like 30mm (see e.g. the Goalkeeper CIWS for a similar concept using the same 30mm gun as the A-10 aircraft). It's a question of hitting the target with a usefully high probability. At longer ranges dispersion is higher, so even if the aim is correct there's a chance it won't hit. And think of how it works; it shoots out a stream of shells, and tracks the target and the shells with radar. But what if the target is doing some evasive maneuvers, which AFAIU many anti-ship missiles do in their final phases? The shells fly out at about 900 m/s, and a subsonic missile flies at maybe 300 m/s. So by the time the shells whizz by the missile it has moved a quite considerable distance, and while the radar tracking can apply the correct amount of lead (with a Kalman filter or similar to take into account measurement noise), if the missile is doing some unpredictable maneuvers that might not help. With the missile closing at 300 m/s, you don't have very many cycles of shoot-see-where-the-shell-goes-adjust-aim before the missile slams into the ship. And to make matters worse, there are also supersonic anti-ship missiles. Now evasive maneuvers in the high-density air at the surface isn't really a concept that works out for Mach 3 missiles, but, well, the missile compensates by flying as fast as the cannon shells. Hence stuff like this RAM thingy is seen as a better system against high-end anti-ship missiles.