Non-pilot here.
Noted in the report in section 1.8.4 about flight recorders is that the investigators were able to go back to a previous flight (on January 12) by the accident AC, with a different crew, that also took a circling approach from the north into RW12. That was 4 days and 23 sectors previously! A cumulative flight time according to FR24 of 695 minutes, averaging 30 minutes per sector.
That's 11 & 1/2 hours of data. Totally possible since the installed FDR had a 25-hour recording time. In an aircraft built in 2007.
Compare that to the hand-wringing on here about the recent runway incursions and near-CFIT incident aircraft having their 2-hour FDRs overwritten by the time the respective flights arrived at destination.
On a similar note I trawled through all the Yeti KTM to PKR flights since the new airport opened on January 1. The accident AC successfully flew this sector 22 times, obviously with different crews, and only twice before made a circling approach to RW12 from the north. The rest were all straight in to RW30. According to FR24 those 2 tracks each had a weird significant RH zigzag about 2/3 the way along the flightpath, possibly a result of missing data points.
Of all the numerous Yeti Airlines KTM to PKR flights I found one circling approach to RW12 from the north that had all of it's flightpath in green (meaning minimal missing data?) - flown by ATR-72 tail number 9N-ALN performing YT671 on January 4. The following screencaps show the final stage of that flight, which has not been shown in this thread previously. The second image is zoomed to show a little more detail.
Source flightradar24
Source flightradar24