Gravar-mail: On the consistency of orthology relationships