Recently and rapidly diversifying taxa remain a significant challenge to phylogenetic reconstruction. In such groups, genetic data from the nuclear genome generally contains little phylogenetic signal, often creating reliance on more rapidly evolving mitochondrial data and/or concatenated approaches for phylogenetic reconstructions. For a variety of reasons, the reliance on mitochondrial data and simplified phylogenetic analyses may produce deeply misleading results, although the extent to which phylogenetic reconstructions are affected is rarely tested. We used data on the Drosophila melanogaster clade to test the influence of concatenation methods, cytonuclear discordance, and assumptions about nucleotide evolution on phylogenetic reconstructions. Our surprising results demonstrate how easy it can be to generate a well-supported phylogenetic tree that may significantly differ from a species tree in a rapidly diversifying taxa. We demonstrate that even when using large volumes of genetic data, it must still be carefully curated prior to use in phylogenetic analysis.