Historically the 'figure of merit' definition used was the ratio of the equivalent thickness of wrought iron penetrated to the thickness of armor plate tested. So if a 12 in KC plate had the same resistance as 28 in of wrought iron, the figure of merit was 28/12 = 2.33. This definition seems quite sensible and intuitive. As is well known, for KC the figure of merit was thickness dependent. It was highest for medium thickness plates. In particular, the figure of merit falls off steadily for thick plates as thickness increased further. This was due to the lesser control of heating and cooling the interior of very thick plates.
Nathan Okun when he developed his armor penetration theories decided on a very different definition of relative quality factors. His definition involves shells attacking 8 in thick plates at normal impact. Late period US armor is assigned the quality 1.00 and all others are relative to this. As Mr Okun acknowledges choosing a particular thickness brings in scaling. Scaling is the characteristic that thick plates exhibit progressively less resistance against shells of increasingly higher calibres. For KC armor, scaling increases as the % unhardened back layer thickness decreases. KC plates up to 8 in thick (of a given era) with very thick faces/low BLT were the most effective, while thick plates with (up to a point) thinner faces/high BLT were superior against large calibre shells.
The figures Mr Okun gives on these pages are as follows:
US Q = 1.00, BLT = 45%
Italian Q = 0.98, BLT = 55% (?)
German Q = 0.96, BLT = 59%
British Q = 0.928, BLT = 70%
French (assumed) Q = 1.00, BLT = 65%
Japanese was not comparable as it was made basically to a WW1 era specification.
Comment: With the exception of the French there is a correlation of Q with BLT. As the US plate has the thickest face/thinnest back layer, other factors being equal, it had the highest Q under the 8 in plate quality criterion. As the British plate had the thinnest face/thickest back layer, it had the lowest Q under this definition. The French plate is the exception and for it to have had Q = 1.00 and BLT = 65% it would have had to be of higher real quality than all the other plates. However, these values are only assumptions and apparently German tests of French plates found them to be inferior due to higher impurity contents.
Mr Okun also gives an armor projectile damage factor. He says this QD is usually the same as the Q factor. It was the hardened face of the plate that generally damages the shell at normal impact. It is plausible that damage increases with increasing face layer thickness. As Q (as defined here) also increases with face layer thickness, it is plausible to put QD = Q.
I have always imagined that the armor quality factors given on this site are primarily the result of real quality in metallurgy: low impurities, optimized element selection in the alloy used, careful temperature control, etc. However, I have now come to the (provisional) conclusion that for armor of a given era the Q values are primarily artefacts of the rather arbitrary definition of quality used.