Although air disasters may seem dismayingly routine to those watching the news, the vast majority of them (94%) occur within a few minutes of takeoff or landing, often victims of low-level wind shear, controlled flight into terrain, or other near-surface hazards. Whatever happened to Air France 447 as it cruised over the Atlantic at 35,000 feet must have been very different. As of this writing, a 3-mile long field of floating debris has been found near the presumed crash site, but otherwise frustratingly little is known. This sort of case is not common--one has to go back to 1948-49 to find cases where scheduled airline flights simply disappeared at sea like this; and in those days planes flew much lower and were more vulnerable than is the case with today's hardware.
All we know (mainly from a few scraps of telemetry data) is that the airplane entered an area of convective activity and experienced some turbulence; that some unspecified electrical or mechanical problem may have developed; that cabin decompression seems to have occurred; and that whatever happened happened quickly enough that no distress transmission could effectively be made. The size of the debris field suggests that the plane may have come apart in the air before impact (which would explain the decompression.) At the very least, something must have happened that caused a sudden and total loss of flight control, making a managed ditching impossible. Nothing else seems consistent with the facts.
So what could do this to an A330 with a presumably robust airframe and multiply redundant electrical and computer systems? Not simply being struck by lightning--that happens to airliners frequently, usually with little or no effect. Could the plane have been torn apart by storm winds seven miles over the ocean? It takes a lot to break the airframe of an airliner; superviolent thunderstorms capable of destroying an airliner outright are known, but mainly over land, almost never in mid-ocean; and the crew was amply equipped with gear that would make deviating around individual cells a routine task. Too, there was nothing even as organized as a TD in the area, just random ITCZ activity. Simply saying that the airliner flew into a thunderstorm falls short of being a satisfying explanation. However, if something caused a complete loss of control, the airplane could have entered a spiral dive whose g-forces could easily have exceeded airframe limits and caused failure and breakup. The question thus hinges on what that crucial something is. Could the weather issue be a red herring, a mere misleading coincidence? Did some hidden airworthiness issue, such as a fatal software glitch, play an unsuspected part? If the flight data and voice recorders are recovered from Davy Jones's locker in the next few days, we may be treated to a definitive answer--or, as with the Columbia disaster or the disappearances of 1948-49, we may merely be left to speculate forever. Time will tell.