From a human factors point of view this is an interesting one. There obviously was a RNAV approach available because the controller offered it. Not having the plate is the tricky bit as with one it would clearly have been the safest option. Nobody wants to hold their hand up and say they are incompetent to fly an NDB, but raw data in a 737 it is far from a regular procedure. A bit of finger trouble in the FMC with a under confident FO and it is surprising that so few of these incidents occur. In the good old, bad old days I suspect some people would have accepted the RNAV approach programmed it in the box (or as an ILS overlay) and without a chart used the NDB minimum. Given the cloud base it would not have been an issue. I am not for a moment suggesting this is the right thing to do. But a weak crew with a lack of situational awareness would have been better off than trying to fly this in vertical speed and heading select. Building a distance ring around the runway at 7 miles or suchlike would also have helped with determining distance to height. I strongly suspect that the ILS out of service was only properly understood just before top of descent and that with a bit more notice they would have been better prepared.