I guess I was just a bit impatient (I figured it out)

In case any of you want to know:

The script was returning a header "Content-Type: text/html; charset=ISO-8859-1", which wasn't being compressed, but a normal .htm document with a "Content-Type: text\html" was being compressed. The line "mod_gzip_item_include mime ^text/html$" meant that anything with the mime type text/html was going to be compressed, but the cgi script also appended the charset stuff to the end, so that's what threw it off.

So I fixed it by using

Code:
mod_gzip_item_include mime ^text/html.*
instead of

Code:
mod_gzip_item_include mime ^text/html$
(So anything beginning with text/html would be compressed.)

Thanks anyway everyone
-Mike