If this is always a 1-1 correspondence between these, with them both representing the same entity, then it would make more sense to keep all the properties on a single node, and just have the application logic select the appropriate fields from the node depending on the level of detail.
The tradeoff here of course is that you wouldn't be able to tell at a glance from the db itself which fields are at what level of view, that logic lives at the application layer itself. On the plus side, you only need a single node for an entity, and you have flexibility in the query itself to decide which fields to include at which level of detail without having to alter the db.