I was checking out Paul Grimwoods site for what's new and came across an interesting point in his Flashover Q&A http://www.firetactics.com/FLASHOVER-Q&A.htm
Question 10 reads:

10. Is it true that RFP or 'explosions' may occur in compartments some way detached from the main fire compartment?
Back in '85 I was at a LODD fire where this occured. A member had entered the apartment/floor above from the fire escape to search. Heavy heat and smoke filled the upper floor with no fire. Soon after entering the window the nozzle team began to hit the fire below. My assumption is water hit the fire apartment window and it failed. Fire lapped up the side of the building and the moment it reached the open window above, the entire room blew.
This was neither a classic backdraft or flashover as the ignition was explosive and the fire was sustained.

The typical backdraft comes when both heat and fuel are present and oxygen is then added. In this case the fuel and oxygen were present but the heat of ignition was not. The smoke was heated to the point of "boiling" but still below ignition teperature. The momentary lick of flame was all that was needed for ignition.

Also, the typical flashover from TRF was not present. The contents had reached the point of giving off fuel due to convected heat, not radiated heat. Hence, after the backdraft/smoke explosion, a full contents fire ensued.

Thanks again Paul Grimwood. Your stuff always keeps me thinking.
You might want to avoid the top lighting in your photos though