Both/all subs are connected to the same output in Anti-Mode, so they act in unison, also during calibration.
Creating individual corrections isn't useful. You really want to calibrate the subs as a system, because this will create the smoothest response. The nulls of one sub usually do not coincide with the nulls of the other sub, so they can be compensated better. All that matters is the overall response at the listening position, not how many subs are used or where they are located.
Individual delays would be beneficial in some cases, but should not be that critical. After all, the sound has always bounced multiple times in the room when you sense it, so what sounds should you match is a good question (because the first bounce has higher amplitude than the direct sound).