Inertial navigation failure, causing a drift in the platform? Perhaps we are drifting away from the ping data which is the subject of this thread. The ping data however interpreted shows the path that was followed. The reasons for that path are a different discussion.
Hmmm, Consideering that the logical option for the satcom installation was to obtain required navigation data from already existing equipment rather than using an independent source, a failed ADIRU (meaning ground speed not available to SDU) would preclude correct frequency adjustments on aircraft transmissions to compensate for the Doppler effect. That may be the reason why the satcom link failed from 17:07 to 18:25 i.e. when the A/C track was close enough to de direction of the satellite and the frequency used was possibly out of satellite RX tolerance.
As I understand it, main consequence on computing possible tracks would be to consideer BFO data as not directly related to true radial speed relative to Inmarsat 3-F1.