I'll second Chiroptera's suggestion, minus his follow-up question:
Chiroptera writes:
Personally, I would like to know whether any method for identifying ID has been tested on objects or systems that are already known to have been intelligently designed or produced without intelligent intervention, and of complexity similar to the biological systems that are under discussion.
This alone is plenty and cuts to the core of Dembski's specified complexity and information theory claims.
--Percy