Logging the max acceleration in x,y,z would have been better than throwing a bear out of plane. Knowing a bear survives is ridiculous and they knew it. Ethics are orthogonal, cadavers would have ben a better choice.
What can you compare that acceleration data to? Put a human on a shaker table and measure his brain damage? At some point you have to damage humans or animals to find out their limits.
You would now have numerical values to compare rather the MILSPEC 'dead_bear' units. Nowhere did I make an ethical argument against animal trials or say that we should do experimentation on humans. But launching bears out of an ejection seat was even stupid then.