I miss two important options for saving linked pages. The most important one is a "stay on site" option which ignores links to other sites (e.g. if you capture www.example.com it wouldn't save links on www.somewhere.else). I also think it should be the default.
The other one is "relative links only", i.e. that only links to lower levels be followed, but not to the upper ones. For example, if I save www.example.com/foo/bar/index.html it should not save pages like www.example.com/foo/index.html even if linked from the first one.
Without these options, much junk is saved if links is followed. It could be corrected by fumbling with file checkboxes (where is "select/unselect all", btw?) and filters, but it requires much work which would be unnecessary should the two mentioned options exist.
I'm afraid I'll still be downloading sites with wget until the time when this, easier solution, becomes powerful enough to replace it. It would be worth five stars from me then.