``There is nothing wrong with a good illusion as long as one does not claim it is reality.''
Howard Pattee ([Pattee 88] p.74)
In the previous two chapters, the behaviour of Cosmos has been analysed in considerable detail. However, as suggested in Section 6.8, it is still unclear exactly what the scientific value of these results is, in terms of what they can tell us about biological evolution, or about a more universal view of biology. It has been hard to escape the nagging feeling that many of the more interesting results we have seen have been dependent on rather specific features of the system's design. On top of this, even though considerable time and effort has been spent in analysing the system, we have only really scratched the surface of exploring the parameter space.
Over the course of this work, I have come to see a number of problems with the general approach that has been taken, if it is to be used as a scientific framework. In this chapter, I will discuss these problems and related issues, and suggest, in some fairly general ways, how the approach might usefully be modified and extended.