Not unreasonable really, within an hour of the crash it was reported that the crew made a transmission that referred to cargo shift. That transmission has not been denied.
But not confirmed, either, which is much more to the point...
If I remember correctly it was the original eyewitness who posted about this, who was on the apron at the time (?) - without disrespect to him, unless he overheard it being transmitted on frequency, then under the circumstances with the natural hysteria surrounding an event like this there would be dozens of conversations about it very quickly and it wouldn't take long for a statement like "the guy in the tower said he thought the crew were struggling with load shift" to easily end up being miscommunicated as "the guy in the tower said that the crew said they were struggling with load shift"
deSitter: I commented on the original roll to starboard a few pages back but it got buried under an argument about handbrakes or something, anyway I cant be bothered scrolling back through all the BS to try to find it! Somebody else commented on it before me and nobody picked up on it then either. I agree that it is an interesting thing to consider. It's very easy to disregard it - especially as you can only notice it in the moments while your brain is still going "oh look, an aeroplane" and there really isn't enough visual footage of that roll present to let it fit into one's interpretation of the events that are going on - so the load shift idea is far more compelling and the gut reaction of anybody with a grasp of flight dynamics is to conclude what everyone is saying: load shift leads to high pitch angle, then high alpha, then unrecoverable stall. This does seem like the most logical explanation for what happened and people have even resorted to the air-crash investigation expertise of William of Ockham who lived in the 12th century to defend the idea.
What interests me about the initial starboard roll, is that if you follow the sequence through to the eventual departure at TOC the motion of the aircraft in the roll axis seems phugoid, which, if that was the case, would imply an instability in roll.
It's very tenuous for several reasons: whatever is occurring in roll we assume the crew are trying to correct it, whilst at the same time possibly dealing with something else (so could merely be a side effect of that); there isn't enough video to reveal the period of the first roll; any upset in roll would be reacted to by the crew, so therefore the timings would be altered; the dashcam camera introduces its own aspect parallax as mentioned by somebody else on here.
Despite the lack of information from the video to make this idea as compelling as the load shift view, if you watch what is going on in roll closely, from that initial recovery from starboard, then to port then back to starboard at the stall, the timing of what is occurring does fit in with an instability issue.
In other words we shouldn't rule out the possibility this aircraft may have had no issues with loading at all, but may have had an unexpected instability in roll from the moment it got airborne. This could also distract the PF from calling for gear up, and with the aircraft committing most of its energy to the climb at V2+10 continued instability in the roll is going to gradually erode the forward speed in the climb, especially if the instability is absolute, leading to greater roll deviations each time, and greater forward speed deviation, exacerbating the issue into the stall.
I can hear people shouting "But what about the pitch angle" - I'm not convinced about that either. It is not as simple as people think to work this out using trig, trust me, give me a camera and I can make any one of you guys look thin. The fact is we think its a steep climb angle because it stalls and we think this must obviously be due to climb angle, which must obviously be due to load, right? By this point we already have Human Factors "Confirmation Bias" going on. Next time you are at an airport with some time to kill - go stand about a mile upwind of the departure runway and watch them go over you, then go do the same from the downwind end and watch them climb away. Regardless of your experience, without the lateral impression of climb angle these angles give you, see how often you catch yourself thinking "Wow that's steep!".
If you watched the same video and the aircraft continued off into the sunset without any drama, you wouldn't even have noticed what you thought was too high a climb angle, because your mind would have automatically compensated for the lack of information in the angle of the shot.
That's enough psychology, but food for thought.