We use that particular token to identify each individual TokenStore instance’s storage. So two TokenStore instances that share a storage will now be recognized as such, and give less false positives on duplicate token claim detection.
From the PR that introduced the feature:
This identifier allows the processors to identify whether two TokenStore instances (either on the same node or different nodes) are pointing at the same location. This is useful for AxonServer when trying to balance the number of active threads on nodes. This identifier can distinguish different processors when they carry the same name, but point at different token stores.