To follow up:

After trying a variety of options from the manual, I gave up and tried wget.

The following achieved the desired result and kept the local sites file
structure, verbatim, between scrapes.

wget -x -m -k <http://website>