I looked into the docs for 'External depth' and 
noticed this (I think it would explain the problems 
some of us are having getting 1 level of external 
sites with the images of those pages):

---------------------------
"Maximum external depth 
Define how deep will the engine seek in external 
sites, or on addresses that were forbidden.
Normally, HTTrack will not go on external sites by 
default (except if authorized by filters), and will 
avoid addresses forbidden by filters. You can override 
this behaviour, and tell the engine to catch N levels 
of "external" sites.
Note: Use this option with great care, as it is 
overriding all other options (filters and default 
engine limiter) 
Note: This option is not filled by default, so the 
depth is equal to zero. "
---------------------------

I think the key to the problem is this "as it is 
overriding all other options (filters and default 
engine limiter)".
It seems the external link filters are applied after 
all other filters, effectively overriding the manually 
entered +*.gif and "Get non-HTML files related to 
link"  Maybe this ought to be changed so they don't 
override anything, and are overrideable.

Oh, last thing, I mis-typed in my previous 
posting..."Get non-HTML files..." is on the Links tab.