"We couldn't say what direction it had gone in, but the plane wasn't standing still because the signals were getting longer, i.e. further in distance from our satellite."
Inmarsat Senior Vice President Chris McLaughlin
That's an odd choice of a word. I assume that by longer he means the time delays between each hourly ping, in milliseconds. If so, it is wrong to construe that delay with distance alone. Ping time can increase for multiple reasons not related to distance. One obvious confounding factor is interference of some type.
What would be interesting to do is some type of statistical smoothing. For example, if the plane was flying a consistent speed one would expect the increasing ping delay to follow a consistent, not random, pattern. OTOH if the ping delays were increasing at an increasing rate then distance alone might not explain it. The forensic calculations get complicated quickly having to take into account so many parameters. I'm going to assume that they go it right and have had many different eyeballs look at it. But I do not think we should treat the result of such calculations as a certainty, more like an educated guess.