A high level overview of what it is that was explained to me:
It’s ‘working time’ vs ‘Idle time’ for a process (in percent, so 0 – 100). Actually, each process has a task list (in the database – that’s the MQ* tables). Processes take a look at the task list and if there’s something to do (e.g. activate a job), that’s ‘working time’ (including database requests, network, …). If the task list is empty, it’ll wait for a task – that’s ‘Idle time’