Ideally you line everything up. However, due to differences in the various signals, it's not usually possible to achieve anything close to sample alignment. If you are able to line up within 1 or 2 msec, you'd be happy.
Given that framework and the window of uncertainty, I am suggesting you put the SBD at the beginning of that window.
Another thing to consider is that you're likely to rely more heavily on SBD if the hall or your location in the hall is bad. That might mean the AUD mics are picking up first reflections that are nearly coincident with the direct sound. That's bad; it can sound like mush. In that case, you might consider purposely pushing (or keeping) the SBD in front of the AUD and turning down the AUD a bit as well.
Your ears will ignore any reflections that arrive a few msec after the direct sound, and reflections that arrive after 40msec or so will sound like pleasant reverb. It's those reflections in between that you have to worry about.
I think there are different approaches to this.
What I've often done is put the aud/mics a few ms *before* the soundboard. My idea is that, even with added soundboard, it should still sound like an aud/mics recording. Just a better one. If I hear soundboard only, it sounds unnatural to my ears. I want it to sound like it was really recorded in the room. I've also tried adding reverb to the soundboard, to make it sound like it was in the room. Then mixing this with the aud/mics.
In summary, I'm using the soundboard for its clarity (no pa distortion, higher vocal levels). But I want it to still sound like a plausible aud/mics recording.
Richard