Apcalc

[Mobhack]

Javelins are top-attack missiles, so they get more than the average top armour hits - therefore the front armour is usually irrelevant. The thin roof armour is what gets attacked. Also note what I said above - HEAT can (like solid shot) also over-penetrate as well as having fuse failures (0 or so pen). penetration is a bell-curve distribution, statistically. Bigger HEAT (and AP) will do overall due to size (WH) than little ones, and little ones like 40mm grenades, are penalised since they may make a little hole and blow out the ear-wax of the crew, but unless the teeny jet hits something vital, they are firecrackers. Bigger bang (WH) is better, especially for chemical energy rounds.

[Kiwikkiwik]

This time I went through the entire German OOB about 74 Heat Weapons. Same pattern.
For every weapon except 1 the APCALC Heat penetration value is LESS than the Mobhack value. For 1 is is the same. If the diffence between Mobhack and APCALC were random then wouldn't you expect at least 1 APCALC value to be MORE than the Mobhack value in a sample set of 74? it never is.
The one exception is the weapon 83 flamenwerfer, but weapon 108 Flamenwerfer is like the rest one less than Mobhack, this may give you a hint as to what the discrepency is caused by.
Also interesting is that for 10.5 cm and above the difference becomes 2 less in APCALC than Mobhack. A couple of very large warheads consistently got less 3 in APCALC than Mobhack.