I run a minimalist arch system running openbox without a desktop environments but i use select components from multiple desktop environments some of which spawn background processes such as kdeinit4 which do not properly quit when i shutdown my X server (they terminate due to an abrupt loss of connection to the X server is what i can glean from the jumbled up output to my tty).
Can a session manager accomplish that? I did not really know what a session manager was from the description initially i thought it might be like a ssh session but reading Wikipedia and Google searches I've become confused. The line i see most is that session managers can save the state of a desktop and restore it later. But i can not figure out if it will just take a snapshot of my current desktop session as i exit X and then just allow X to deal or not deal with the running processes or would a session manager be able to detect running process started during the X session and close them without terminating them abruptly.

Am i barking up the wrong tree with this path and if I am is there an elegant solution to the problem or would I need to create some sort of hackish script if i wanted this ability.

It is up to the applications to decide what to do when the X server quits.

Session management is a hard task - remember that the processes are organized into a tree, and that the "session manager" is not necessarily at the top of the tree (at a minimum it is one level down, and there is one or more sibling trees involved). That means that terminating a session SHOULD be done by telling the "session manager" you want to log out. The session manager can then send an event to known processes that want such an event, then they wait for a reply. After a timeout, the session manager may then terminate - and terminate the X server. Many such processes now (gnome anyway) use Dbus for such notifications. If the application doesn't use dbus, then it cannot register with the session manager, OR request notification of a logout request.

Note - all of the above is strictly voluntary. If the X server terminates (or is terminated), no such voluntary event message is ever sent, nor is a response waited for. It is up to the application to choose what to do when the socket connection is closed. Well written applications will catch such events, and make suitable arrangements about terminating. Unfortunately, not all of them do; nor do they all even exit (they seem to keep trying to reconnect...)