Am I missing a clue to how mounts work?

On a standard Linux system, filesystems listed in /etc/fstab are mounted by a startup script, which runs mount -a. On an empeg there is no startup script; init and the player just mount the filesystems they know about. The listings in /etc/fstab are just there to keep tools such as df happy.

Mounting the root filesystem (read-only) is done by the kernel before init is started (err, obviously; the kernel couldn't run init otherwise!).

If you want /dev/hda2 mounted as /usr/local automatically, you'll have to do it yourself in a non-Empeg-supplied init script.

Peter