The previous thread is found here:
Amish's spreadsheet takes into account shield regen, hull repair, armor, and shield mitigation. I found it hard to play with, because it appears to be hand calculated, but maybe it's just because OO Calc can't open the file correctly. I don't think JJ posted his spreadsheet yet, but his post seem to indicate that it was the case.
Not sure what the conclusions are exactly, but here are the problems encountered thus far:
1. Neglecting all reduction and regeneration, assailant should win by ~20s. (JJ, opening post). Does not predict the right winner.
2. Taking into account hull repair, shield regen, armor, and shield mitigation; and adding a 2s delay to mitigation decrease, illum wins by ~150 health (Amish, reply #7). Does not predict the winner shield mitigation numbers.
3. Taking into account hull repair, shield regen, and shield mitigation only, assailant should win by ~20s (with_an_e, reply #9). Does not predict the right winner.
The test appear to indicate that shield mitigation is based on damage prior to shield mitigation and armor rather than the reduced number.
It's possible that the developers don't really want the players to know the precise game mechanics, possibly to discourage long theorycraft discussions or to reserve the ability to make tweaks to the system without making large chunks of user knowledge obsolete.