On the Uniqueness and Complexity of Reconstructing Repeat-Annotated Phylogenetic Trees