I'm having the same problem.

From what I can tell, the "capture URL" feature is allowing you to log in FOR
THAT "CONNECTION SESSION" ONLY.  What I mean by that is if you disconnect from
the site you're mirroring (not under your control, HTTrack seems to be doing
this), the login is dropped.

I surmised this is the case because I can download part of my target site (a
distance education uni site) but then I get oodles and oodles of login pages
for all the rest of my links.  Most frustrating.

Of course, I could be wrong.  If anyone out there has better information on
this, please let us know.