OMP is intended for roughly the same tasks as OJS, but for books / monographs instead of journals. Specifically, just like OJS, it's intended to help manage the submissions right from their start with the author, through the review and editorial processes, to prepare them for publication. Once that is done, it can also present them for readers (for sale or for open access publication) on a public-facing catalog and feed data to external services like Amazon through the ONIX data standard.
It does not provide an editor for the text, i.e. to actually write the content; that is done through external programs like word processors.
OJS is mature, whereas OMP is still under active development and not yet released for production use. As such, there are many parts of OMP that still need some work. It also does not have as much breadth as OJS yet for things like plugins, integration with other applications, etc.; those details will come with time.
OMP is also a testing ground for a number of user interface features that will eventually make their way into OJS to modernize it -- AJAX, JQuery, modals, grids, and more.
The software roadmaps on the wiki might also be helpful; see the various roadmaps available at http://pkp.sfu.ca/wiki/index.php/Main_Page#Software_Project_Pages
Public Knowledge Project Team