I don't know for sure the advantages of doing them separately, unless you have a work team doing separate parts.

I keep it all together, because I don't like any sub-renders if I can help it. For instance, I have 21 tracks stacked with mostly audio tracks, but I also have 4 camera tracks with audio in the bunch too.

Sometimes the separate audio doesn't line up with the video or one camera is out of sync. I normally stack everything together. Then see which ones can be lined up and be sync'd without stretching or shrinking. Go to the end and make sure one track or the other hasn't drifted off sync from the beginning to the end. They may all line up without drifting apart, but not always. Sometimes you can 'Zoom' in on the timeline and sync the audio by looking at it in the timeline. You can match the peaks and valleys, this is also an easy way to see if they stayed in sync at the end of the tracks. If not in sync, you have to stretch or shrink the tracks to whichever one you want to be the 'Base' track. (Hover mouse at end of track in timeline, when a square Icon appears, press and hold Ctrl, a wavy line will appear under the square, still holding Ctrl, pull the end to stretch longer or push the end to make shorter) This is not to make the tracks the same size, but to adjust for a slight difference in the speed of the recording so they can all match and the pitch is the same.
Once the audio is matched from all and everything is sync'd, you can shut the audio off to the cameras and just use your audiosource or you can use volume envelopes on all tracks and pick and choose which audio you hear predominately.
I'll stop there, as it may be too confusing to give too many details at once