Despite the poorly phrased and possibly trolling nature of the now locked penetration thread, I do agree that the penetration modelling seems a little quirky.
That's probably because penetration depends on so many variables.
An units effective armor is probably somewhere between 1-1,5 times the listed point value. This is to reflect that hits do not come at a straight angle, and in fact armor is sloped for that purpose. The actual armor that protects is calculated for each shot individually: shots do not arrive at the same angle. There's some sort of a critical hit chance, but I've got no idea how often those occur or what the model for them might be. Since firing units have some scatter, some shots will hit on the side if the target is at an angle.
Penetration is probably 2-2,5x the listed value at point blank range. With a bad armor roll and good penetration roll light weapons will do damage: this is sort of made worse by the rate of fire that autocannons have (although I think penetration is calculated for each individual burst, not each individual round). Your formula is most likely very close to the truth.
re: morale, until we get a table with all variables that effect it we'll just have to guess. I feel the morale effects are spot on. Without seeing replays it's hard to judge what exactly caused the unit to crack. Low skill, lucky hits, volume of fire, penetration, platoons of one tank, death of allies? Who knows? Please include a replay if you think something is amiss.