With the advent of uninative (and it being enabled by default in poky), NATIVELSBSTRING isn't really the right variable to be triggering a rerun of the sanity.bbclass tests on since it's always the same when uninative is active. We need to be using a different variable that actually represents the host distro.
The sanity code already looks at lsb_distro_identifier() instead of ${NATIVELSBSTRING}, so I'm not sure if this is a valid bug.
Closing as the code appears to do this already. Paul, please re-open if there's a detail I didn't consider.