It would also be helpful to know how the toroid was mounted, and if the failure occurred where pressure was applied. Toroids are usually mounted either flat with washers and rubber gaskets with a single screw through the hole, or vertically in an "Omega" bracket, with some rubber cushioning material around the periphery. But in either case there may be additional pressure on a "high spot" where two windings may be crossed, and softening of the insulation from overheating may also contribute to a short.
It is fairly easy to check for dielectric breakdown between the bifilar windings, but nearly impossible to detect a weakness between adjacent turns of a single winding. But it might be possible to use a higher frequency, or PWM pulses of higher peak voltage, to produce a higher potential between turns of the same winding, and observe spikes of current due to breakdown.
Paul