policy-node: schedule a rescan without timeout if defined target is not found
This MR schedules a rescan (which happens after core sync) if the defined target of a node cannot be found, instead of scheduling a rescan after 2 seconds. This seems to solve random pulseaudio application hanging issues if user has a complex configuration with lots of nodes.
Fixes #146 (closed)
Might also fix #67 (closed) and #102