Vertical it ain't.
The attitude of the aircraft can be determined by anyone who needs to do so, based on the geometry of the video. The video provides an adequate gnomon, the wing span, to ascertain the body length to span, and also using the wing span and aspect, to evaluate the height above the ground. Additionally, the frame rate is around 20FPS, or 50ms, and so time to traverse from TOC to impact can be ascertained. It is steep... and hits hard. As the tip vortex is also observed, an approximation of attitude to FPA is available which can be converted to angle of attack. Do the trig. The height above ground also gives the FOV of the video, and the distance from the observer. The acoustic signature of the post impact rapid fuel burning is evident on the video sound track, and that gives a good measure of the distance from impact of the observer, using the time delay and local speed of sound, neglect speed of light.
The fact that this aircraft had already operated a sector with the primary load in place makes it more curious as to causation. Report is going to be worth reading, WRT how the aircraft came to be in such a condition. The wing fuel loading would move the cg aft, but it is not a massive shift in the scheme of things, the B747's CG general envelope is only to 32% basic from memory, and out of trim case in the past have gone back past 44%, at higher weights, i.e., IU's much further along.
veers left, then right wing slalls.
Teldo serious, #342
Doubt that there is enough evidence available to support your assumptions at this time. Unless you are picking up the rudder image pixel change from rudder aspect change, then that is s bit early to say. The wing drop to the right may be aerodynamic, it also may be controlled by the flight crew, it is not inappropriate in itself, in fact broadly follows TBC and AI guidance for departure form controlled flight. The rate of roll is near the aerodynamic limit for aileron only, it is not anywhere near the limit of the rudder secondary effect authority on roll rate. Generally, in large aircraft upsets, the roll off at TOC, if any, is not particularly pronounced, unfortunately there is a fair amount of data out there to support that opinion. As often as not, the aircraft maintains a relatively stable roll attitude. Why? the normal forces at the low speed case are not that high, and so the roll moments that may exist are not high, but the aircraft maintains a high inertial moment... not much roll.
[For Capt Bloggs..."at TOC"] Did the aircraft stall? very likely, as it evidently had a pitch up moment that was not able to be countered, and eventually the flight path has started to decay. At that stage the KIAS is low, well below normal Vs1g, as the aircraft is not maintaining a level flight path, even though it is accelerated by the pitch up rate. Assume a 0.2g pitch acceleration which is pretty agressive for a B747, as the aircraft goes through 45ANU, the stall speed is going to be below the level flight stall speed, (at 90ANU, it is only related to the pitch acceleration i.e., 0.2g... if the rate is 0g at 90ANU, the aircraft doesn't stall, and the flight path is purely dependent on thrust relative to mass.g...). Did the aircraft roll of due to stall, not necessarily. Would a crew response to roll the aircraft at TOC be appropriate, Absolutely. Will it help? depends on the situation, but it is better than assuming nothing will work.
[For Capt Bloggs..."on the descent"] the aircraft is probably stalled, given the angle between the tip vortice and the aircraft longitudinal axis. Stall is related to IAS for a g loading, it is related to angle of attack, so g loading is a primary factor. 0g, zero "stall".
The crew are going from RPT pilots to test pilots in a few seconds, trying anything to maintain a reasonable attitude is not only the best choice, it pretty much is the only choice, unless you have a really good religion. The later attitude is consistent with a flight crew fighting to survive a catastrophic problem. The bank angle they have attained gets the nose down rapidly, but there is inadequate blue sky to effect any meaningful angle of attack reduction before impact. Less roll angle would not have resulted in the nose getting down as far as it did, and would have been ineffective as well. These guys tried hard, and should be respected for that. With more favorable conditions perhaps it would have been only a harrowing hangar story. On the day, they did what they had to in those last seconds, professionally.