Refactor node manager to only run the publishing loop on the highest level nodes #87
Description
As mentioned in #86, our current infrastructure introduces too many nodes on the system, clogging our communications with redundant data.
This issue is to refactor the node manager to only publish data on the highest levels of the node hierarchy (ex. a subsystem object instead of individual components) while allowing the member nodes to still be accessible (similar to having private variables in Python).