Is there anything to confirm uncommanded reverser deployment other than the pictures showing the #2 sleeves deployed (only have observed outer sleeve in pictures) and some unconfirmed accounts?
I have been studying the CFM56-7b thrust reverser (78-34-00) design to understand the barriers for uncommanded reverser deployment. it is designed to prevent this failure from occurring that which has no single pathway (assuming the throttle levers were not physically commanded to reverse thrust). A cascade of failures including multiple independent switch-controlled relays would be necessary for both sync-locks to activate, for both locking actuators to release, and to apply hydraulic power to deploy. No single hardware failure would lead to deployment (the lever/linkage itself is the only common element). There are maintenance procedures to manually deploy and override locks. The logic for on-ground or 10 feet AGL is applied as an additional interlock to the hydraulic commands.
The thrust reverser is not efficient in comparison to forward thrust, especially at low speeds (e.g. effective reverse thrust would be reduced from comparable forward thrust). For sure, there would be no forward thrust when the sleeves deploy the blocking doors (when considering yawing moment).
Until 60 knots, there should be considerable scrutiny on setting takeoff power. With the data suggesting this speed was never achieved, then the pilot monitoring should have had full attention on the engine instruments, and should have noted reverser deployment without any delay. I would have expected both throttles to be slammed to idle in this occurrence, in less than three seconds to allow for reaction, and engines spooled down about three seconds later. That would suggest that the airplane should stop accelerating within about 6 seconds from the failure occurrence. In this case, the considerable loss of directional control left the airplane still accelerating while plowing away.
Two points from the ADS-B reporting:
1) there was at least a minute from when the airplane reversed course in taxi and lined up to the when the airplane began its takeoff roll.
2) the airplane continued to accelerate after departing the runway. Three specific reports were six seconds apart, the first at runway departure (and it about six seconds from initiating the takeoff roll). From departing the runway, peak speed was reported six seconds later and speed was subsiding 12 seconds later.