Thank you both - two quite different, but equally valuable responses.
The accident is consistent with pilot distraction insofar as it occurred just after his observer had conned him around the base turn onto the attack heading. Lt Kershaw confirmed that he had acquired the target. The aircraft struck the water a few seconds later. My one reservation concerns time and distance, against height lost.
The aircraft passed over the finals marker at 500 feet descending, according to the observer's altimeter. By my rough calculation, at 420kts, it would have taken around 13 seconds to fly 1.5nm. That sounds like a lot of height to lose, and a long time to be distracted from height monitoring. Assuming a constant rate of descent, the aircraft would have reached 150 feet - release height for laydown - at just over 1nm. The run in to ship target 3 was 9nm; there was no need to be that low so soon.
The electrical evidence suggests there may have been a problem with the rad alt.
A speculative scenario, very much up for discussion: aircraft is flying the downwind leg of the range circuit with rad alt left on the 500ft scale. aircraft passes the finals marker; observer calls the height. Pilot expects the rad alt to start reading on the 500ft scale at any moment, at which point he will begin to use the display needle to control rate of descent to run-in height. This system works time and again. The pilot is focused on the tasks of lining up on the attack heading and acquiring the target. The rad alt display is offering no reading and no 'Off' flag. This does not cause great concern, because there is confidence that it will read imminently. The pilot is distracted from how long he has allowed the aircraft to continue at the rate of descent from when he completed the final turn. The error may have been partly one of time perception.
Is the electrical system controlling weapon release in close proximity to the any part of the rad alt system? On the previous laydown attack, their 4lb practice bombs had failed to release from a CBLS on the port inner pylon. I wondered if the two events could be connected.