Quote from
mm43:
"ROP = 023°T x 4.34NM(5 miles/8km) from LKP
Those with a calculator can come up with their own position, but the one given above will lead you to the main engines."
One assumes that both engines would have detached in the first second after sea-level impact, and then descended practically vertically the 3900m to their present position on an apparently horizontal ocean floor. To all intents and purposes, I think we can treat the sea-level impact position as approximating the engines' PPOS.
It seems implausible that a jet airliner, particularly one starting from cruise altitude, could average little more than a statute-mile-a-minute of GS ("ground" speed), even if it was in a deep-stall condition throughout the period. This would be somewhat slower than a Citroën "Deux-Chevaux".
It's unclear to this writer how the flight control system and PF could have allowed the AoA to progress above the conventional stall, even after the disconnection of the AP and downgrade to Pitch-Alternate Law at around LKP. Also, Alpha-Floor would have selected TOGA thrust before the stall, adding further energy. One has to assume that the aerodynamics of this conventional aeroplane, even at cruise CG, would also provide protection from deep stall. Others may like to calculate the TAS at Vs-min for various altitudes.
Allowing for the priority given to position reports in the ACARS-transmission hierarchy, the latest possible time at LKP was roughly 02:10:20z, and the earliest possible sea-level time was about 02:14:30z. At the end of this period of at least 4 mins10 secs, the aircraft seems to have impacted the sea only about 5 statute-miles, (26400 ft) away in azimuth. Assuming direct line of flight in azimuth, average GS would be 72 mph (62.5 kts). Average wind component during the descent is unlikely to have contributed: there is no suggestion of significant headwind.
My tentative conclusion is that direct routing in azimuth from LKP to sea-level impact does not provide sufficient mileage for the minimum flight time of 4 mins 10 secs. A series of tight S-turns might be possible, or a spin; but some kind of R/H or L/H teardrop, commencing after 02:11z, seems more likely, as I've proposed
before. The teardrop would be unlikely, of course, to have been a tidy one.