Nagios Fusion exists to offer a "single pane of glass" for multiple Nagios Core instances. It doesn't interface directly with Thruk, but if the original Nagios Core CGIs are still reachable, Fusion should work fine.
Assuming you wanted the entire state of all zones to be held by a Nagios Core master, passive checks are generally the answer to that. You'd essentially rig up your individual Nagios Core zones with something like an
NRDP client to ship their results periodically (or in real-time) to an NRDP server located on the Nagios Core master. You'd still need some sort of clever configuration management to wrangle all the individual zones'
object definitions and join them together on the Nagios Core master, though. Else the passive checks wouldn't have the corresponding object definitions for each zone. Sounds like a fun
Chef project

I don't know of any simple step-by-step instructions for such a setup, though I've definitely seen it done with larger
Nagios XI setups acting as master for several Nagios Core instances (replacing a config management utility like Chef with the Nagios XI API, essentially).