Here is one example of a small piece of fuselage that was peeled off during a break up process. Peeling continued rearwards until structure was weak enough to break apart close to wings, not close to tail.
I wonder if aerodynamic drag could have been large enough to yaw the plane significantly right after initial event when there was an initial hole on the left side, close to wings? I don't know how autopilot behaves if it is not able to compensate that kind of yawing? Switches off? Alternate law activates? Could pilot then break VS off by too large rudder movement?
Then I haven't seen any picture of floor. It seems that seats were shedded off the floor one by one. Could that mean that floor is also in small pieces?
It also looks like not all seats have scorching marks from fire. I remember that doctors said, scorched were passengers from very rear part.
It seems to me that after initial event on left side, that side started to shed pieces of fuselage and floor increasing aerodynamic drag and making the plane to yaw and change direction. It looks like initial event created hole between cargo hold and passenger cabin. Shedding continued until planes structure became too weak and it broke off close to wings while aerodynamic drag bent tail sharply to right, breaking it off from the remaining fuselage.