I'll start by saying that this language came from an earlier version of the notebook, and I think it is fine to change it to circle. My guess is that it was originally written this way because we are querying the Hierarchical Triangular Mesh resolution, and so when we specify level=10, we are really specifying the size of the triangles, not the bounding circle.
Section 2.3.1 is one of the places where DM feedback would be useful. What I think is that most astronomers will want to get their hands on a FITS file, but the DM standpoint has been that you'll never need the FITS file. I'll follow up with you on Slack.
Questions for reviewer @kadrlica :