Oh geez, the anti-ship missile is not aiming for the barrel of the CIWS where all the bullets spit out. In that case, it would be heading straight for the bullets! The stream of lead from a CIWS can cover only one range of vectors that would hit the ship... others would miss entirely.
Besides, by the time a mach 3 missile is within 1 or 2 kilometers, it's momentum will punch a hole in the ship anyway, regardless of whether or not it's "intercepted" by a CIWS.
These geometry of these engagements in not what you imagine. A missile like Exocet or Moskit is aimed at the ship's beam where it has a good radar return, so things like CIWS or RAM are shooting more or less head on at the incoming round. Not preceisely but close enough. CIWS can definitey follow a missile that is corkscrewing or making the climb for a terminal dive. RAM has more than sufficient control authority to ourmaneuver a corkscrewing supersonic missile like Coyote and shoot it down. We do this in exercises.
Regarding the debris punching holes in ships, this is simply not true. In training exercises we shoot down target drones such as Coyote at mach 2.8 that is programmed to fly all the expected terminal maneuvers we know about that threat systems like Moskit fly. Coyote is pretty close in size and mass to a Moskit, actually flies faster and lower, and mimics Moskit in every way. We shoot them down in training exercises using live rounds of Standard, RAM and CIWS. Real all up ordinance is fired and these drones are blown to bits. No debris touches the ship. How that idea gained currency is beyond me. We also have Chukar to mimic things like C-802, MM-40 and certain subsonic versions of Klub with great fidelity. We shoot these down too. Debris hitting the ship is a non problem, it does not happen.
Missiles like RAM do not have to impact their target, that is why they carry proximity fuses of some discription. A near miss is actually what they require, so the rods in the warhead can chop into the missile.
In fleet testing RAM hits a supersonic target on the first try over 95% of the time. We've shot over 150 of these things and that is the actual success rate. Usual practice is to fire two missiles at each target to guarantee a kill. You guys debate as if the USN never actually fires a live round at a supersonic missile. We do, we know the geometry of these engagements and our systems are well designed to defeat known threats. There is a RIMPAC exercise going on in the Pacific right now, and at least 50 Chukar's are being expended, shot down by US and allied warships in these engagements. This is typical of a RIMPAC exercise, they occur semi-annually. Hundreds of drones are expended each year for training and weapons testing.
In any event, most of these missiles will be spoofed by the SLQ-32 and Nulka long before they enter the engagement envelope for RAM or CIWS.