chud
I believe that I gave the reasoning for it having been on in the critical phase - work through it again and it should be logical to you. Whether it was switched off after impact or somehow got like that due to impact or for whatever reason, it had been functioning and, as I have argued, had to have been used at least up to the position of waypoint change and the stored GPS data thereafter was consistent with the system continuing normal processing - as Tuc asks, it would be interesting to know if switching off just cuts off the CDU or stops all processing - but it does not matter.
If you did your own detailed chartwork using just GPS data from the nav data (as opposed to getting mixed up with the Doppler) I think you would see that it all fits.
Pulse
I thought that myself at the time and asked her - she said that it was not uncommon for some of them to have digs off the station - no matter, it was just what I was told by the landlady of one of the bigger B&B establishments - I was staying there and it just came up in conversation about the crash.
The bottom line is that detailed analysis of all the known data so strongly suggested a controlled approach to a known HLS that I predicted, with the difficulties that the weather gave for identifying a particular point on the land and judging closing range to it, that such a crew would only have closed in confidently with a PRC112 marking the spot - the fit was only confirmed subsequent to the analysis and so this is a classic case of QED - good proof.
At the very least, a scenario worth investigating fully - however uncomfortable to confront.