There are two different approaches on this subject: the submix is created by an engineer on the (main) mixing console- this scenario allows creating full stems (incl.dynamics, effects etc.). The second option is that the Personal Monitor Mixer system receives a 1:1 of the FOH console channel and on each unit a submix is created.
While the second approach, having an individual submix available at each mixer might sound better to an engineer in the first run, think about it twice: who will do submixes from 40 channels? Usually not the musicians, so now it is back to the engineer. A big disadvantage versus creating one submix on the console is the fact that it has to be done remotely (hopefully with a configuration tool) and for each and every mix station! That adds time.
Creating submixes on the console is not only time saving but also allows much better quality . Most consoles these days allow creating full blown stems utilizing dynamics, EQs and effects, while providing much more consistency through a performance (sudden changes get adjusted in the global submix and the “problem” is solved fast and quick for everybody… ).
The major argument we hear against a global submix is something like: “but the drummer wants a different drum submix then the vocalists”. Of course! And it’s not a problem: give the drummer “kick, snare, hi-hat individually and create a submix (mono, if you have the channel count stereo) for the tom and overhead mics… For “non drummers” make a submix/stem with all drum channels . So creating submixes/stems needs a close look as to who needs what. You might find that 16, 24, even 32 channels can easily be gone…. So it’s clearly an advantage if your PMM system has a high channel count and allows for easy expansions.