Compared to the original test, the above graph had 2 changes.
First, I changed the usb c cable, since the results I was getting originally looked like it was possible that the data transfer was messed up.
Secondly, I removed the EPA foam mat, and put the print directly on the shelf.
I saw on another forum post (Input Shaper issues? weirdness - #11 by SuperCow) that someone had a similar looking graph to my most recent, with lots of low frequency amplitude. In this post the solution was to get the linear rails running more smoothly, so that is going to be my next thing to change. I’ll post the results once done.