It’s possible to make a reasonable 3 megapixel camera without autofocus – see the Sony Ericsson TM717 (Equinox). So what’s wrong here?
I think some of this can be fixed with a firmware update, but I also suspect that the lens is optimized for portraits (i.e. it can only focus on subjects about 2-3 feet away) .
The Devour fares better than the CLIQ for video recording, with 640×480 pixels (VGA resolution) at about 24 fps.
Here is a sample video (download original):