Hello? Pedro , yes the occurrence depends on both there are some values used for modelisation that can lead to that like the minimum dynamic leaf gap if i remember well, and the maximum speed entered for the machine not the one for your plan.
I noticed when I do a plan with a spacing of 2° it is almost systematic to have the RED MU , but when doing a plan with a 4° spacing ,more than half of the plans are ok without red MU.
I'm using 16.4 both manually and Evolution spacing 2° grid 0.3cm , on more than 300 plans, I NEVER had the red MU , unless you play with the prescription afterwards. like ( normalizing to 95% etc etc) and with the 16.4 it show on wich Control point you have the "violation" of speed gantry rate ( deg/sec ) or the leaf speed (cm/sec).
One of them is colored in red if your prescription modification violates some constraints.
I gonna make the devil's advocate here , i think one the reason some bugs or glitches have not been solved til these days , is because the optimizer was not their "property" and the "just" coded around it , and now the made their own optimizer they can solve some issues like
In autoplan no matter with minimum segment area you asked ... you will have a final plan with some CP area below this value .. not anymore in 16.4.
Same with the maximum delivery time .; now if you ask 90 it's gonna be 90.
Doing a plan with auto planning the Modulation of the plan is some kind of surprise at the end .. nor anymore you can choose the lvl of freedom for modulation and MU etc etc
And more i'm not even talking of the new speed... doing a pelvis with node in 16.2 and in 16.4 ..
I Think you should try 16.4 and see.
Hope i helped a little.