Turning multiple source clips into a synchronized, composited (teaching) video

I have been using OBS to create composited videos in realtime which include multiple sources. One of the things that I have been doing is using the OBS Move Transition to animate the entry and exit of the various sources (see 4s example on the left and 40s example on the right, below)  
I have used OBS and the Source Recorder plugin, along with some external devices, to capture a number of different sources and angles (e.g. screen; face-on camera; side camera; tablet) and have been sure that each includes audio.
My question is what would be a good workflow to replicate the kind of realtime compositing that the OBS Move Transition allows using only stock Final Cut Pro X? For example after viewing the realtime compositing, I might decide that it would have been better to stay with the slides (or switch entirely to my camera) instead of showing both simultaneously. It won't be possible to make that kind of change from the OBS-composited output.
Thinking about the problem abstractly, the following workflow seems to make sense:

1. Synchronize the various sources into some kind of combined entity
2. Edit the combined entity "linearly" (in the sense of removing unwanted parts)
3. Split the edited-and-combined entity back into perfectly synchronized edited streams
4. Composite the various edited streams to create the animated transition effects

A Multicam clip seems to be the best way to handle (1) especially since I was careful to include the audio in each source and FCPX seems to synchronize on those pretty well. "Compound Clips" might also be relevant but it doesn't seem like it.
As an aside, a quick search for how to edit the sync points in an exiting Multicam clip doesn't reveal anything useful; it seems that you have to set your markers perfectly before you create the Multicam clip
(2) also seems pretty straightforward: drag the Multicam clip into the timeline and use the Blade All tool to chop and then delete subsections as necessary
What I can't figure out is how to do (3) and (4) assuming that they are the correct things to do.
So my question is essentially what's the "best" way turn a set of source audio and video files into the kind of composited video that OBS can do in realtime?