Follow up to this!
I received a LaserBee-A from someone who wanted me to add my v3 datalogging board to it.
With the exact same laser, untouched since the initial tests I got a reading within 2mW of the initial test on my Ophir.
This was what I got on the LaserBee-A.
LB-A 1.810W
Ophir 1.844W
LB-II 1.656W
Unlike LaserBee has suggested it now appears that the LB-II was reading incorrectly, and my Ophir head is reading correctly. That or both my Ophir, and the LB-A are reading incorrectly, which I highly doubt.
I would still like to know if the LB-II's are curve adjusted individually as I believe that's what was the result of the LB-II inaccuracy.
Either way, the LB-II is inaccurate or the LB-A is inaccurate.