That and not too much damage in the first place. Not all hit the flight deck and not all who did penetrated it. Coming in at a shallow angle could result in the plane sort of sliding across the deck. but I´m sure Cal Bear knows more about these things.
Furthermore the open hangar and the armoured hangar deck often limited damage to the uncritical areas above the hangar deck.
The best features of the American carrier, at least related to battle damage, were the wooden flight decks and the open hanger space ABOVE the armored deck. An American carrier, especially later in the war, could take a serious hit and be back conducting air operations in a few hours. This wasn't always the case, witness the
Franklin, but it often was. The deck was quick to repair, which was why the U.S. preferred the wood to metal, and the open hanger design, although a pain in really heavy seas, allowed blast effect to vent outward (compared to ships, espedially the Japanese carriers, with tightly enclosed hangers that allowed maximum blast effect AND trapped fuel and oil vapor).
As far as the kamakazi itself, the biggest problem was that most of them couldn't really fly. They'd had basic flight school, but hitting a 30 knot target thats doing its level best to avoid being hit takes some skills. There were, as you point out, a LOT of glancing blows, where the plane either literally skipped off the deck, and near misses where they just flat missed the ship. Since most of them didn't carry really large bombs (many of the trainers couldn't get airborne with more than a 100kg weapon as payload) near misses didn't do as much damage as an actual near miss by a bomb, which would generally be 225 or 550 kg.
Still, there were LOTS of kamakazis and they would have been a much larger problem except for several American tactical adjustments:
1) Increasing the number of fighters carried, sometimes by 100% from pre-war levels. This wasn't a bad as it initially appears, since by 1945 the fighters could carry as large of a bombload as many dive bombers.
2) Radar picket destroyers. These poor brave bastards are the mostly unsung heros of the kamakazi battles. They were posted as much as fifty miles out from the main fleet, along the threat axis to provide an additional half hour or so warning to the fleet. Of course, being the first Yankee ship the kamakazis saw, and since most of the pilots couldn't tell a destroyer from a battleship, it could get very interesting to be the radar picket. There are documented cases of
A (singular) picket being attacked by 30-40 planes, especially once the Japanese figured out their mission. It sucks to be a pawn in a chess match.
3) This last one is less of a tactic than a technological breakthrough. The proximity fuse made a fairly effective weapon, the 5"/45 into a real killer, by spring of '45 they were even introducing 40mm shells with the damned things. Absolutely shredded the attacking aircraft.
Kamakazis would have been a problem, especially against troop transports, but in the fairly short term they would have been wiped out. Once they started to fly, the hidden airstrips would have been located, with the resulting bomber attacks and fighter groups lurking nearby night and day. By the time the invasion started, most of the 8th, 9th & 15th Air Forces would have been in theater, with all the fighters that had sawed the Luftwaffe off at the knees flying out of Okinawa. That would have put around 1,000 first line fighters (mostly P-51 & P-47) within easy flight range of the Kyushu bases. That would have been in addition to the 1,000 or so Hellcats and Corsairs on Mitscher's fast carriers and an additional 700-800 on the CVE force.