That is what baffled me at first until this clue. The actual locking device is armed to engage or disengage with the lever along with the throttle block. However, in some form or fashion mechanically or hydraulically the elevator mechanism has to be separated from the yoke and commanded down for the lock to engage. So obviously the falling forward of the yoke realigns the yoke or doing control checks accomplish the same result of resetting the yoke to the elevator commands. If the pilots were holding neutral position of the yoke, released the gust lock on taxi and then the lock itself would disengage, BUT did not do control check then the yoke would not reset itself to align with the elevator control commands. So by falling out of normal start sequence they missed both opportunities to realign/reset the elevator functions. That is my limited interpretation anyway. VFD
It's not quite as complicated as that, there is no mechanical or hydraulic "separation" of the yoke from the elevator. Pulling the control lock lever upward merely causes a latching pin to slide into a matching hole in one of the bellcranks which are part of the normal elevator mechanical control interconnection from the flight deck to the aft equipment bay. Pushing the control lock lever down causes the latching pin to retract. The latching pins for elevator, ailerons and rudder are all controlled by cables which have a common connection at the lock lever.
The elevator's natural position on the ground with no hydraulic power applied is fully down, which corresponds to the yoke being fully forward. The position where the locking latch will engage is very close to the full down position. When engaging the lock, the pilot might have to pull the yoke back an inch or two until the lock lever will come fully up, but it is not a significant distance.
The preliminary report indicates that the locking latch for the elevator was found disengaged, and though it could have retracted from the impact forces of the crash, I would think that if it HAD been engaged at impact, there would have been some mechanical damage to the latch, or it's mounting bracket - apparently there was no indication of that.
From a psychological perspective, I would think that if the yoke was locked, or mechanically jammed in such a way that the yoke could not be pulled back - that the pilot's instinctive reaction would have been to instantly recognize that a serious problem existed, and to have initiated an abort immediately.
Instead, all indications are that the abort was only initiated many seconds after VR, which seems to point to the possibility that the yoke came back normally, and the problem was not recognized until the nose failed to rise as would be expected in a typical rotation, consuming many precious seconds, and many hundreds of feet of remaining runway.
If so, this would point to a failure in the mechanical control path between yoke and elevator in which the elevator did not respond to movement of the yoke.... but in such a way that the yoke travel was not restricted. A broken control cable would be my first thought - but there are other possibilities.