[StarBaseSweeper]

Ok I re-did calculation.
There is however two different hypothesis:
H1: When a 5kt damage hit a component but is not enough to destroy it, the 5kt damage will be added to the next shoot, whose target will be randomly chosen (not necessarly the same one).

H2: When a 5kt damage hit a component but is not enough to destroy it, the 5kt will be added to the next shoot, which will target the same component, if it is targettable (no X Only weapon).

For me H1 is wrong because it makes the armor used in adamant mod useless:
(for 1 armor of 30kt)
Hit 1 5kt, hit armor.
Hit 2 5kt+5kt=10kt, hit armor.
Hit 3 5kt+10kt=15kt, hit component => destroy 1 5kt component, 10kt left, hit armor.
Hit 4 5kt+10kt=15kt, hit component => destroy 1 5kt component, 10kt left, hit component => destroy 1 5kt component, 5kt left, hit armor.
Hit 5 5kt+5kt=10kt ....

With H2, we have:
Hit 1 5kt, hit component, 1 5kt component destroy
Hit 2 5kt, hit armor.
Hit 3 5kt+5kt, still armor.
Hit 4 5kt+10kt, still armor.
Hit 5 5kt+15kt, still armor.
Hit 6 5kt+20kt, still armor.
Hit 7 5kt+25kt, destroy armor.

Anyway using both hypothesis, with teh following configuration ships:
Components: 10 5kt components.
Ship A: 10 5kt armor. (total 50kt=> 50% of ship)
Ship B: 1 30kt armor. (total 30kt=> 37% of ship)

Shoot of 5kt.
For 4 shoots, the leaking damage is:
Ship A: 10kt (on total of 20kt damage)
Ship B, H1: 14.8kt
Ship B, H2: 6.6kt

So given the result of SJ test, I would favor H2 as being true.

Of course this for only 4 shoots (but the calculation is very complicated for Ship B,H1 because of reported damage and new targets)

Any confirmation for the hypothesis?