By the time the USN was fully involved at Okinawa, much had been learned about the methodology of the Kamikaze attacks. Kamikaes usually worked in groups of 5, and not all of the aircraft were suicide planes. Typically, the flight consisted of 2 aircraft with the more experienced pilots acting as escorts for the other 3, whose task it was to crash into the ships. The 2 escorts would generally try to draw the CAP away from the other 3 planes.
There were two preferred methods of attack, often determined by the weather over the target. The flight would come in at high altitude, generally over the CAP, using cloud cover to mask their precise location, then drop down onto the targets.
The other method was to come in low under the radar, get as close as possible, then climb up to 300-400 feet and do a shallow dive into the target.
This, of course, changed the way the CAP operated, and actually led to the requirements for a point defense fighter, that ended up being the F8F Bearcat and F2G Corsair. The gap was bridged by having F6Fs fly high CAP, and by keeping F4Us on the flight deck ready to launch, since the F4Us had better climb rates. The enemy flights could then be picked up by radar pickets and the appropriate CAP vectored in.
Coordination of the AA defenses were based on target range, and although the CICs controled the overall battles, it was generally left to the gunnery officers to determine which targets where a.) within range b.) posed the greatest threat. Individual ships' CICs were often informed when targets appeared on the scopes of other ships, but out of range of their own (remembering the distances that some of these groups covered, and the many quirks of the SK and SC radars.
Typically, the 5in DP mounts would open up at extreme range and fill the sky with flak if they got past the CAP. These were controlled by the Mk 14 Directional and MK 22 height finding radars. If the aircraft got through those two lines of defense, then the 40mm would open up, each mount being aimed by a Mk 52 Fire Control Director, and later on, by Mk 56 (?) radar controlled directors. Then, if it got real close, the 20mm guns would open up, with a gunnery officer controlling a battery or gallery of guns. Usually, when things got real heavy, gunners knew what to do and needed little guidance, especially by 1945, after the initial shock of the Kamikaze wore off and tactics were developed to counter them.
As far as air recovery operations were concerned, if an aircaft was in need of landing while the carrier was under attack, they were usually routed to another carrier nearby that wasn't under attack. Remember, the Kamikazes didn't operate in the huge waves of aircraft as often thought. Typically, although waves of 40-60 were experienced, groups were usually small, with one or two attack groups, 5-10 aircraft, taking part during a single attack.
Although it was possible to try and recover an aircraft during an attack, it was not standard practice, as the AA gunners would shoot at anything that got too close, the ships were generally maneuvering, often radically, and they weren't always headed into the wind.
That's my two pence, but it's very late and I can't ever remember my own name at this point... Don't quote me on the Mk types of the radars either, I'm pretty sure those are the numbers, but I'm just too lazy to get out of this chair to confirm.
Jeff