Window 2:
Top Pane is for Post Processing Additional (downloads segments of the release to find audio/video/nfo/par2). This pane collections release information and stores it for future script usage in the database. Once PP Add runs, NFO processing runs next. If an NFO wasn't found during PP Add, it will look within the NZB for the specific .NFO posting and try to grab and import it into the NFO table.
Middle Pane is for Non-Amazon processing but is really just for TV/Movie/Anime post processing. By post-processing I mean we directly tie a release to third party information (such as the TVDB ID for TV, the IMDB ID for a Movie, and the AniDB ID for Anime). This helps us organize releases by their respective content in Cover Views.
Bottom Pane is for Amazon type processing (Console games, Books, and Music) but also includes Games and XXX processing. Again, the feature of this pane is to tie releases to third party IDs and retain their information locally for display and processing use.
Window 3:
Yes, that's exactly what this pane does. It connects to our IRC channel on SynIRC and listens for PreDB notifications from our collector bot. When it sees a new post, it inserts the Pre into your local PreDB database. During insertion, it also calculates hash values for the Pre title for hash matching. If the Pre includes Request ID information it will also insert that as well. This pane is vital for making good renames as quickly as possible, though it is possible (but not recommended) to not run IRCScraper and just import the daily dumps.
Window 4:
This pane is there because you turned on the show processlist option within the Tmux settings. For reference, this pane will not work unless you have a ~/.my.cnf file for the user running Tmux specifying all the needed MySQL login information.