The "Why"
According to the time stamps in HazelNuts39's excellent graph, all hell broke loose prior to and during the minute of 02:10, with fifteen messages time stamped during those 60 seconds. By message 16, beginning with time stamp 02:11, the situation had likely degraded to irrecoverable, and by time stamp 02:14 SATCOM sent the grim hint of an aircraft falling from the sky ("Cabin Vertical Speed").
JD-EE's failure analysis trees might be characterized as "event trees," and they are, indeed, complicated and many because the data are simply too sparse to easily "hone in" on the event. Start with AUTO FLT AP OFF and work backwards... What would cause this message? Unusual attitudes? Failure of GEN1 and transient electrical disturbance? Tie in a branch for PROBE-PITOT disagreement causing out-of-parameter airspeed calcs. Tie in a branch for overwhelming meteorology. A branch for radar (failure). etc. etc. Such an event tree is , as JD-EE stated, "very large," but when properly weighted, this upside-down tree might point to the WHY in a best-guess methodology.
An understanding of the A/C's attitude during its fall might help to narrow the search area beneath the 02:10 LKP, a good effort by HazelNuts39. But it seems to me that any answer as to WHY the upset occurred that can be derived from the ACARS messages lies in the group of messages time stamped 02:10. The rest were sent as the plane went down, and again, they don't tell us WHY it went down...
A readable CVR and FDR would "rule in" conclusions in the event tree(s) about why the aircraft was upset. Analysis of a recovered hull and debris might "rule out" other conclusions. Pardon the double negative, but I cannot imagine how BEA could refuse to continue the search! As mm43 counsels, stay tuned for a new phase in the new year.
GB