Well, I'm not a biologist or taxonomist so you can take what I say with a grain of salt. But I've been thinking about this for a while now in terms of object libraries (I'm a software engineer by profession). Modern object libraries are most commonly organized using namespaces based on two criteria:
1) inheritance (class A inherits attributes and behavior of class B)
2) interfaces
Note that class B may inherit and possibly override aspects of A (the class from which it is derived) but may also just use the aspects of A as designed by A. Interfaces roughly correspond to building blocks that are there for all classes to use and may be also be overriden. Although a terminal derived class could be thought of as a "species", it's not since it may usually be derived from at some later date as deemed necessary. That's why class hierarchies use a dotted notation for denoting a class (at any arbitrary level). Interfaces are not really relevant to the definition of a "species" in class libraries. They are just noted in the class definition itself (ala B implements interface X). It seems reasonable to my simple mind that phylogenetic trees could be treated in a similar manner to object-oriented class libraries with some minor modifications.
™ Version 4.2