Question about the difference in satp3 policy detector setup

I'm wondering about this small difference in how satp3 runs the "Detector Setup" block: https://github.com/simonsobs/scheduler/blob/a4173fee02f43316f3036d2aa9380e132ded3a73/src/schedlib/policies/satp3.py#L183-L184

The other policies run this as just:

run.smurf.bias_dets(concurrent=True)

This relies on the defaults built in to the underlying code, which runs with rfrac=(0.3, 0.6) on all bias groups.

I guess my questions are:

Does bias_groups=[0,1,2,3,4,5,6,7,8,9,10,11] represent all bias groups or is that a subset?
Why call with rfrac=0.5 vs the default rfrac=(0.3, 0.6)?

My motivation here is supporting the detector operations fully in sorunlib so that direct access via pysmurfs = run.CLIENTS['smurf'] isn't needed. After #109 is merged this'll be the only remaining direct use of the SMuRF clients. The answers to these questions will inform what arguments we should consider implementing in sorunlib.

simonsobs / scheduler

Question about the difference in satp3 policy detector setup #110