Argon, That's an interesting video. All samples other than the bearing are flattened without visible energy release, probably by uniformly heating the items. But the ball obviously releases significant energy. First, I would calculate the energy stored in the elastic energy of deforming not just the ball, but also the press. You can see the dents in the press plates, and you know the modulus of steel (I assume the plates are some alloy of steel). But just as obviously, the plastic energy needed to deform the plate is not released, and so not added to the energy used to deform the ball.
As a check, you can estimate the Ultimate Tensile Stress of the ball material, estimate the amount of deformation and integrated force to get the entire ball to that state, then estimate the temperature (from the color of the sparks) of the flying pieces, and compare them. If they don't agree within a large factor, there's something else contributing to the total energy.
Another check would be to record the press ram pressure X diameter (if you can get that data) and integrate as a function of the distance traveled. Then subtract the energy that went into deforming the press plates. I'm sure I got some details wrong, but a check on units is always in order.