I would generally do the compressor and eq on the main outputs (L&R) and I also usually use a channel for the effects return (for EQing and such). Otherwise it looks great - and also there is no reason you can't do it the way you have it except I don't often see multiple inserts on the output so you would have to chain them there on the one insert. If you are seeing multiple insert points in the main section, they are probably for the subgroups.