The debris trail from the plane shows conclusively that the vertical stabilizer and rudder came off long before the engines. Once the plane lost the vertical stabilizer it was completely out of control. During that time, it would've experienced some high side loads as the plane yawed uncontrolledly. The engines came off as a result of excessive loads on the pylons while the plane was out of control. From a structural standpoint, it isn't hard to understand why the pylons both failed, considering the plane was at or near full power and was experiencing sideloads that were never anticipated during design. No engineer designs a plane to fly under those conditions simply because things like this are virtually unheard of and are unsurvivable when they do occur.

The key to solving the crash is to find out why the vertical stabilizer failed. The NTSB .pdf file that was at the bottom of the article indicates that at the moment of failure, the plane was flying at 255 knots indicate airspeed. It was experiencing an 8 to 10 degree left sideslip, probably as a result of the turbulence from the 747 wingtip vertices. Apparently, the rudder was fully deflected to the right. Airbus engineers have calculated that under those conditions, the vertical stabilizer would've experienced excessive loads. That might have lead to the vertical stabilizer failure. I speculate that it's possible the vertical stabilizer was weakened as a result of unequal load distribution caused by that composite repair back in 1988.