I have been testing the new b491 version of monowall and zeroing in on the captive portal features along with the traffic shaping features while performing numerous speed tests. In my findings and my reading I have seen that you must have 'per user bandwidth restrictions' turned on in captive portal in order to effectively traffic shape....is this true?
Depends on what you want to do. If you just want to do some overall traffic shaping (e.g. fair distribution of the available upload bandwidth) and not give a fixed (possibly RADIUS-controlled) limit to a particular user, then the "per user bandwidth restrictions" are not needed and you can simply configure the traffic shaper as normal.
My tests having both traffic shaping turned on and set to a download and upload value in comjunction with the per user restriction set to those same values in CP seem to conflict and cause slower speeds. When I turn off per user restrictions in CP and leave just TS turned on it seem to perform better speed tests and more reliable.
The per-user bandwidth restrictions apply on the captive portal interface; it's a bit like dynamically adding pipes and rules for each user at the top of the list (so if you have per-user restrictions on, then other traffic shaper rules on the captive portal interface won't have any effect). However, if you have traffic shaper rules on your WAN interface, then the traffic will effectively be shaped twice (once on WAN, and once by the per-user bandwidth restriction on the CP interface), leading to reduced performance.
I'd recommend against mixing normal traffic shaping rules and per-user limits.