Starting with Julian Lage immediately optimizes the talent factor!
Mounting the omnis under the stage lip essentially turns that pair into dedicated audience/room-facing mics (they'll still provide low frequency reinforcement in either position). That can be good for providing more isolation of that stuff from the direct stage sound, which allows for mixing in more of it when appropriate and is where the level riding / compression of it described above can be best applied, or it can be a not so good thing if you are relying on the omni channels for pickup of some of the on-stage sound as you describe.
Regarding the level riding of that content, either done manually or automated, as described above- I tend to simplify thinking about it to essentially targeting 3 different levels: An increased level of ambient / audience / room content between songs and during quiet passages, the standard amount you'd typically use in a mix if you weren't making further adjustments or automation of it, which probably gets applied to the bulk of the performance made at intermediate SPLs, and a reduced amount whenever things get loud enough to begin "overloading" the reverberant room acoustics, thus lessening the tendency for things to start sounding overly crowded and muddy.
No doubt that's rather advanced mixing strategy. I'm sure some TS members reading this are rolling their eyes. Certainly no need to go that far unless you want to play around with taking things further to see how far you can optimizing a recording made in this way. There are plenty of other things you might optimize without doing this kind of post work. I really only mention it because recording in the way you are doing makes it possible, its a cool thing to be aware of, and it can influence how you think about recording things moving forward.
