@Telkine2018, how about the following:
1. Store processor name as entity setting (like station names, or manipulator settings, or assembler reciepy)
2. When a processor is deployed, check if its content matches content of other processors with the same name
3. If it diverges -- update the deployed processor as if "Import" button was used.
I know, it may cause wierd scenarios, but is at least a workaround for the blueprints updating.
As an alternative to "diverging" condition, could also check if the content is just empty.
Also, it may be an option setting (like "mining as tagged entity") - either flag or behaviour selector.