First, a disclaimer: I don't work in aerospace fields at all though I do design aerodynamically correct lifting body airplanes.... A lot of the conclusions here are my research as a layman. Rather I am writing here to put theories down on paper, as well as weigh relative weights of the theories. Also, my heart goes out to those who have lost loved ones in this tragedy, and if any are reading this, I hope it helps put some of the media speculation in more context.
Now.... What is known: At 0200Z (GMT), the pilot of AF447 sent a manual transmission that they were flying through a storm system. This corolates well with Tim Vasquez's projections and analysis but is way off from the BBC's maps. The plane would have entered the backward edge of the mesoscale convective system (MCS) and would have exited the forward edge, where the storm cells would have been strongest.
At 0210Z, the plane sent a series of ACARS messages denoting a large number of failures from 0210Z through 0214Z. These messages are designed to speed aircraft maintenance rather than determine the cause of an accident so they lack certain details which are important in this case. At the moment, however, they are one of the more important sets of information which is publically known.
After 0214Z, no further details are known. The vertical stabilizer was eventually recovered, but it isn't clear where or when it broke off yet. Most likely the vertical stabilizer was broken off by sideways forces but at the moment it isn't clear whether this happened in flight or during impact.
Finally, we have the Air Comet pilot report where the pilot at 7N49W reported seeing a bright light in the distance following a vertical downward trajectory for six second. Due to the curvature of the earth, the Air Comet pilot was not in a line of sight to the AF flight. The AC pilot could have seen a meteor.
Theories and weighting (note the weightings could change rapidly with new information):
1) Initial messages caused by lightening strike. Probability low to moderate. A lightening strike to the Radome could damage Pitot tube systems, weather radar, inertial reference systems, etc. The lightening strike would have to enter or exit on the radome to cause this sort of damage. Such would seem generally unlikely due to the tolerances involved. The main reason to suspect radome destruction is that pitot tube icing itself can't explain the TCAS fault reported. However, the inertial reference units are near the pitot tubes so it seems to me that severe turbulance-related damage would be more likely. Recovery of nose section, radome cover, etc should be sufficient to eliminate or confirm this possibility.
2) Meteor strike causing destruction of radome. Probability: extremely low. This would have a similar damage profile to the lightening strike scenario if the meteor was small enough to avoid further damage but large enough to destroy the radome. Also unlike a lightening strike, these are not events which frequently happen. Recovery of radome cover should be abe to rule this out or confirm it.
3) Pitot tube icing resulting in unsafe speed of aircraft. Probability: Moderate to high. Pitot tubes are known to ice up in conditions where no liquid water exists. For example a 1999 meteorological flight reported ice and graupel from 18k feet upwards through 41k ft and the DC8 involved in the 39-41k ft. range reported pitot tube icing. This would suggest that pitot icing can occur from processes different from structural icing. That case is worth reading in comparison to the present tragedy because it is reasonable to see the storms in both cases as comparable (both were equatorial meso-scale convective systems). The problem though is that the TCAS (Terrain Collision Avoidance System) faults might not be explained by simple cases of pitot icing because that system relies on groundspeed and GPS measurements rather than airspeed indicators. However, if severe turbulance was encountered (perhaps exacerbated by the autopilot increasing thrust to compensate for low airspeed readings), this might be sufficient to cause damage to aircraft systems including the TCAS and the IR systems (more on that below-- note though that the IR Disagree errors occur the next minute suggesting that they probably occur after the TCAS fault). While this seems like the most likely explanation, barring additional evidence to the contrary, it isn't possible yet to suggest that this is entirely certain. All of the 0210Z messages, however, except the TCAS error could be explained by the computer recognizing bad input from the Pitot tubes.
After the initial incident, the ACARS messages paint a picture of rapid deterioration of the situation. At least one internal reference unit failed, and shortly thereafter both primary and secondary air control systems would have failed. It is unclear at that point whether the aircraft was in direct law, or on manual backup (which gives LIMITED use of the rudder and elevator trim). The manual backup systems of an airbus are not designed for turbulance or even landing (they are only designed to provide some troubleshooting time while a plane is in-flight).
The next question is whether the aircraft broke up on impact or whether it broke up in the air. At the moment, there does not seem to be sufficient information to say. The last message, indicating a fault with the pressurization system due to external pressure increases COULD indicate decompression at that point, but it could also be due to cascades of bad information from the Air Data unit or an actual increase of outside pressure due to a rapid descent (for example, after a mach tuck allowed to progress too far due to lack of inertial reference). While it is likely that more detailed analysis of the vertical stabilizer will help answer this question, it is too soon to say whether it disintegrated in the air or when hitting the water (or a mixture of both). (The vertical stabilizer appears to have been broken off bit sideways force but whether this was the result of a sideways crash or problems in the air is currently uncertain.)
All in all, this is my rating of hypotheses surrounding the crash.
At this point, the evidence is not sufficient to conclude much beyond this IMO. Unfortunately a lot of this has been the subject of wild speculation from the media. Such speculation probably does not help anyone who is in search of truth whether due to curiosity or loss. I hope my post helps clarify at least one layman's view of the evidence for any such folk.