Lots of good points, DB, but it seems to me that there were a lot of other factors contributing to the crew's apparent loss of SA. For example, a) nobody is at their sharpest at 2:00am, b) they flew a recognised company route that I presume was intended for use in poor weather and, surely, nobody would approve or continue to accept such a route that overflies a 280 ft rock, c) for some reason it seems possible that they believed the rocks to be low, and so were happy to overfly them, d) they were unfamiliar with the local topography. Points c) and d) may have reinforced their reliance on point b).
Forgive me if you know the next bit, but here goes: the aircraft's roll channel would have been coupled to the nav kit to follow the route. When initially alerted to the need for a heading change the PF followed the usual multi-pilot procedure by asking the PM to select the roll channel into 'heading' mode so that the PF could then manually select the heading bug, and hence the aircraft, to the desired heading. That's absolutely fine and safe in normal fully-coupled circumstances but, in this case, neither pilot appears to have appreciated the urgency. Instead of overriding the autopilot by manually moving the controls to change the aircraft's flightpath rapidly, as would probably have been appropriate, the pilots used up valuable time by following automation SOPs. By the time the urgency was emphasised it was too late to avoid the island.
It seems that the pilots were complying with the tech crewman's direction, but in an inappropriate (with hindsight) manner. SOP techniques in non-SOP circumstances.