Monitor number of stale services

Thanks for the research! Since the existing OMD checks already do livestatus queries (I guess), this shouldn’t be too hard to implement.