I don't follow this at all.
Normally, pallets distribute their load through the rollers. The rails are horizontal guides only, and do not directly carry any vertical loads. In fact, since the normal cargo nets are attached to the edges of the pallets, the edges would be more likely to be warped slightly UPWARD, and NOT transfer any down force to the edge of the pallets.
After re-reading my posts I can see the misunderstanding. I will clarify. The pallet rails Boeing installs provide full three dimensional restraint: vertical, lateral, and longitudinal. The downward vertical load is distributed to the floor beams through the seat tracks by way of the rails AND the rollers (the rails and rollers are attached to the seat tracks.) The rollers pick up most of the downward load which enable the pallets to be be moved around inside the aircraft during loading. Powered rollers assist in moving the pallets laterally and longitudinally.
The load on the pallet is netted down to the pallet at the pallet's edges. Thus the upward vertical loads and the left and right lateral loads are picked up only by the pallet rails. The rails provide all upward, left, and right restraint through the pallet's edges. The forward and aft longitudinal loads are picked up by the pallet locks which are part of the rail system. So the rails also provide all fwd and aft restraint, again thru the pallet edges.
Since this load was (reportedly) a "floating load" the rails and locks picked up none of the loads. The upward vertical loads and left and right lateral and fwd and aft longitudinal loads had to be picked up by the various straps and nets restraining the pallets. Each D ring the straps and nets were attached to were only good for 5 klbs. It seems likely that the restraint scheme was inadequate, the restraints failed, and the load shifted aft. If (as reported) parts that belong in the tail section aft of the pressure bulkhead were found on the runway (presumably on the departure end of the runway beyond the point where the aircraft rotated) then the load shifted aft violently enough to rupture the pressure bulkhead AND the aft fuselage skin, enabling parts to fall out of the tail section after rotation. If that was indeed what happened, it seems likely that there was insufficient pitch authority to counteract the aft CG shift, resulting in the crash. There's a small chance such a failure could rupture the hydraulic lines controlling the elevators, which would result in loss of pitch control to counteract the aft CG shift, also resulting in the crash.
Hope this clarified.