Although this crash bears a significant resembelance to G-REDL, there is a big difference here (to me) in the failure mode. G-REDL report states that:
"The upper section of the gearbox
assembly, which included the remains of the second stage of the epicyclic
reduction gearbox, the conical housing and all three lift struts, had detached
from the MGB and were recovered with the rotor head."
So the second stage epicyclic gearbox AND conical housing had broken free and remained with the rotor head. This being consistent with the finding that the gearbox casing had been burst open, due to the broken planetary gear, and the casing fractured all the way around releasing the rotorhead. They also noted no distortion in the Barbecue plate to suggest suspension bar failure.
In this accident the conical housing does not appear to be connected to the lift bearing. Neither is the reduction gearbox section. We can clearly see the splined connection of the rotor mast suggesting it has slid out of engagement. (There is also a photograph of the rotorhead being airlifted away from the crash site to be examined and the conical housing is not visible here either - ill see if I can find it...).
The splined connection cannot simply "slide out" of engagement unless whatever is holding it in place is released (Suspension Bars/Conical housing to gearbox). IF this were another epicyclic destruction then we would surely see the same failure as in G-REDL, with half the planetary gears or at least the conical housing still attached as the failure of the gearcasing would be at the outer gear ring.
The Barbecue plate also seems very bent upwards on the front edge, which would be consistent with a rear suspension bar failure and rotor disc pitching forward - but could this damage have occured during the impact with land?
The amount of panels, cowlings and air intake screens scattered suggests to me the rotor struck the front of the aircraft rather than the tail boom during the break up.
With this in mind, it could be the reason that AH are so adament this is not a mechanical fault within the gearbox. And I would have to agree on that basis. An epicyclic failure would rupture the gearbox casing and, as in G-REDL, it would still be attached to the rotorhead. In this case the rotor mast has slid out of its splined engagement which suggests whatever holds it in place did not do so.... And if that pin was indeed missing from the suspension bar eye....
Only time will tell.