Currently, if we try to combine a tree chain read from rb and another one from mb, check.chains() throws an error: somehow, only the multiPhylo from mb has $tip.label. Using setequal to compare $tip.label in individual trees, in contrast, returns TRUE.
Could it be that the problem lies in read.nexus() vs. read.tree()? In that only the former adds a global $tip.label to the resulting multiPhylo?..