With the long total duration from 0 to full lock it becomes more important to shave off some script run time where possible.

In that spirit I relaxed the tolerances for ASC convergence in CARM_ON_TR, which is just before the automatic WFS relief:

                # tolerances retried on 20160713
                # wfsTolerancePit = [5,200, 100, 300, 200, 100] 
                # wfsToleranceYaw = [5,200, 100, 300, 200, 100]
                # tolerances from 20160713, intended for speed-up
                wfsTolerancePit = [5,400, 200, 600, 400, 200]
                wfsToleranceYaw = [5,400, 200, 600, 400, 200]
 

If the quality of the relief gets worse, we might have to revisit then. (I suspect that the thermal heating at 40W is the bigger problem though.)