Thank you much, fiddlosopher, for your quick reply. Clearly, I need make use of lua filters to adjust for the HTML code in my Jupyter notebooks. I've now read the lua documentation and tried some of the examples on the pandoc doc site and pandoc/lua-filters repository - but I don't see an example of how to search for and transform HTML-format tables to native pandoc-format tables, or similar such thing. Can you provide me some additional guidance?
Raw HTML in a Markdown document (which is apparently what the R display_html function produces) will not be passed through to PDF. Similarly, raw LaTeX will not be passed through to HTML.
One approach might be to generate *both* formats in your notebook. You could try that.
Another approach would be to use a Lua filter that matches raw HTML blocks and uses pandoc.read to convert them to native pandoc tables, which could then be rendered in PDF. I believe there are examples of such filters either in the Lua filters documentation, or in the pandoc/lua-filters repository, or on this list. If you can't see how to do it, just ask.