I would anti-recommend Purcell, but I acknowledge that for some people it’s the best. It’s more wordy and “tell rather than show” than e.g. Griffiths.
On Reichl’s book, I want to note from what I’ve heard (not personally read) that the 2nd edition has much more explanation and intuition that the 3rd edition cut out. I haven’t read other statistical mechanics books and so can’t compare to others.
I’m surprised to see Sakurai here rather than Griffiths. The latter is the classic undergraduate introduction, which would seem better targeted to this audience. The topics Sakurai has that Griffith’s doesn’t are more technical than any non-physicist is likely to care about (e.g. the Heisenberg representation). Griffiths’ strength is that he “speaks to you”, making it feel like 1-on-1 tutoring rather than a theory paper. I learned from Griffith’s 2nd edition (blue cover), and although the 3rd edition is out now (red cover) its reviews so far seem mixed: ht
...
Only barely related, but Grassmann numbers are hilariously weird. Among other properties, their square is always zero (though they’re generally non-zero).