Wind-Tunnel,
I am a bit bothered too with the data. Here are my last (Montecarlo simulation) results:
The last south legs of all the simulated trajectories are at constant speed (and altitude) and the south turn occurs randomly between 18:28 and 18:40: it is assumed that the C-channel BFO measure of 88 Hz at 18:41 is a valid one (it is included in the observed BFO profile the trajectories are trying to mimick) and suggests a south bearing at 18:41. The constant (ground) speeds span from 330 kts to 530 kts, the crash latitudes from 22°S to 41°S:
https://drive.google.com/file/d/0B3s...it?usp=sharing
https://drive.google.com/file/d/0B3s...it?usp=sharing
The first part of all the trajectories comply with the ADS-B measures (they are not randomized on this first leg), and roughtly with the Butterworth radar track above the Malacca strait (supposedly linked to the MH370 ). When the south turn is triggered, a constant speed reference trajectory is built (intersecting two kinds of small circles on the earth spheroid: the ping rings and the positions reachable from one given location at a given constant speed) from the A/C position and randomized:
https://drive.google.com/file/d/0B3s...it?usp=sharing
The trajectories best fitting to the observed BTO/BFO values are around 460 kts (true gound speed) and end near 36°S, not near 30°S (ATSB):
https://drive.google.com/file/d/0B3s...it?usp=sharing
https://drive.google.com/file/d/0B3s...it?usp=sharing
the BFO errors versus the crash latitude:
https://drive.google.com/file/d/0B3s...it?usp=sharing
the BFO errors versus handshake instants and ground speed:
https://drive.google.com/file/d/0B3s...it?usp=sharing
the BFO and the BTO enveloppes of the 50 000 flights:
https://drive.google.com/file/d/0B3s...it?usp=sharing
But we do not know clearly what was the injection/turning point toward the last south leg in the ATSB report.
Jeff
PS) Rerunning the simulation (16h run time) to confirm these results.