I would like to understand EXACTLY how the highlighted value below interact to enable Transmission->Bandwidth Control policies.

Specifically, is it by exceeding a percentage of EITHER the UL or DL, Both added together, ???

Since I am mostly constrained by UL, how would I configure it to only engage when the UL exceeds a percentage (ie set DL value very high?)

What kind of hysteresis is built into the control algorithm, ie Control is triggered, but the policies take the BW below the trigger point, how do we avoid flapping?

I've noticed that the traffic shaping control algorithm allows overshoot (underdamped), ie if the limit is set to 5M, the client can often shoot up to 10M before levelling back off around 5M long term.  Since I'm trying to avoid packet loss, can I assume that 2x is a typical overshoot and thus design my control limits accordingly?

I have 3 SSID/VLANs, let's call them Public/Internal/Security.  I am thinking I would like to stagger the policies such that Limit_Public kicks in at say 70%, Limit_Internal at 90% and Limit_Security at 95%.  Currently it's an all or nothing solution.  Alternatively, could I have tables for each limit rule and set limits from 0-50%, 50-70%, 70-90% and 90%+ based on usage (again back to the either, both or something else question above)?

Appreciate any insight that can be offered!

for completeness here's what the User Guide says...rather ambiguous :)