I'll try to defend myself on the definitional argument first Deluded.  

I don't see how my definitional description entials existence. I understand my definitional argument for 'supernatural' to be viewed as a kind of adjectival concept. If I maintain an adjectival conception of supernatual, without any reference to existence as a necessary property, I don't see how I am entirely crippled on my definitional account by your argument. I agree with what you said about the concept 'supernatural' if you include existence as a property of 'supernatural'. However, I am under the impression that my argument avoids this problematic existence quality. If this is the case, then I don't the difference between a class representing 'unicorns' and a class representing 'supernatural' (In the naturalistic way I've described without reference to existence). 

I'm low on time at the moment, so I will read your article again and take a second look at my referential argument. To be continued. In the meantime, can you point where my failings are in what I've written in reply to your paper, so if necessary, I can also rework the entirety of my position.