The freeze point of the fuel may well have been -57C when tested, but if a kerosene is cooled to wax precipitation, wax drops out and is not redissolved into the fuel, then the freeze point of the Kero will be lower after the event. In other words the freeze point of the fuel would have been above -57C prior to wax precipitation, if it had occured.
However, if that were the case with BA038, the question arises what happened to the wax and why were no traces found? (I presume the investigation has included GC analysis of the bulk fuel and residual fuel on the filters for comparitive paraffin component profiling)