The problem might be how much warning Pearl Harbor receives.
If it's a little much, say 6-8 hours, Kidd will try to sortie his ships, if one of the battleship were to sink in the channel, that would put the base out of action for months. The Japanese also had a couple of subs on patrol just in case a ship or two got out. If it's a couple of days Kimble will sortie the whole Fleet to seek out & engage the IJN. That would be a disaster. The Japanese would tear the old battleships apart without air cover, but no one understands that yet. That means the same ships are sunk, but this time in open ocean, where they can't be salvaged. That means Oldendorf's gun line isn't available for shore bombardment or to stop the Japanese surface force at Leyte Gulf. (Of course with two days warning, the Japanese probably won't launch the strike.)
Now, if it's about three hours, that's ideal. You can get the alert to the fighters, scramble everything so it's not caught on the ground. The ships in the Harbor can get enough steam up so the powered equipment is on line, they can set full watertight status, and survive one or two torpedo hits without sinking (or turning turtle like the Oklahoma).
The results of an attack with the defenses on full alert would be interesting. The U.S. had a lot of P-40's on the Island. They were outclassed by the Zeros, but bombers are meat on the table. It isn't unlikey that the Japanese could lose 1/3 of their airwing (perhaps more). the U.S. at least half of the fighters, easily more. Figure less than half the overall damage that occurred IOTL. No Battleship actually lost (excepting a miracle hit like occurred on the Arizona) and reduced damage to the remaining ships in harbor.
The overall impact on the war is hard to gauge. The Japanese would probably be more cautious & the U.S. is likely to be overconfident. The end will still proably be in mid-late 1945 but it would not be hard to project an earlier end. It seems very likely that the course of the war, starting with Wake, would likly change.