in addition you can consider the following comments:
In terms of characterizing in/out group, i think it's not really just pangenome pipelines, while yes they can do this, but any clustering method can be employed. you can get in/out groups through cgMLST, hqSNPs, kmer based methods.
Is it worthwhile to have your mermaid diagram as a figure? or no because they can just look at the code repo?
some proposed edits.
in addition you can consider the following comments: