Firstly, a thank you to Silsoe Sid, who has put a lot of effort and thought into his many posts, and has shown a great deal of patience in answering questions, many posed not for the first time.
After over 2000 posts, it seems to me that we are arriving at a degree of consensus around there being a small number of potential hypotheses for the fuel situation which arose on the aircraft. In no particular order, these are:
1) Gauges, cautions and warnings worked normally, and a severe case of press-on-itis came over the pilot causing him to try to use his final 10 minutes of flying time after the red lights to return to base. ( personally, I discount this, not least because he would have been totally prepared for a flameout and the ensuing autorotation, and would definitely have been on top of the transfer pump situation.)
2) The fuel contents probes were reading optimistically, and no yellow cautions were shown. The red lights did illuminate, but the pilot considered them to be anomalous, and in the absence of the yellow cautions and the presence of fuel as shown by his gauges, was convinced he had enough fuel to complete the mission. (This is my hypothesis of choice. The main tank showed correctly 76KG, whilst the supply tanks erroneously showed satisfactory readings. He knew the transfer pumps were off, and knew he would need to turn them back on at some point, but also knew they would require management given the relatively low level of fuel in the tank. When the first engine flamed out, he was still convinced his gauges were correct, and that the problem was something else. Turning the prime pumps on and trying to restart the flamed out engine could have seemed logical and a good use of time, and could explain why he would not seem to have been well prepared for the second flameout.)
3) Due to a failure of the CAD, the pilot was denied information regarding the fuel state. ( I discount this on the grounds that the first flameout would have been a very rude awakening, and in the absence of data to mislead him into believing he had adequate fuel, he would have sought to land immediately.
I hope my summary hasn't overlooked any hypothesis which is still at large.
As an aside, while we await more data, is there a view as to why the operation of the transfer pumps is not fully automatic? The processing power required is negligible, and it would seem to offer a useful reduction in pilot workload.