>>We can quibble over details, but it’s pretty obvious there shouldn’t be any unused DNA in a biological system.<<
I think the term ENCODE is using is “few unused sequences” not none.
With regard to programming in Cobol and Fortran, code had to be a lot tighter with such little memory but even then there were comments and inefficiencies.
Best to keep the “comments” in the documentation. Check out this evenings story about nano size superfast memory just developed by some guys in Pennsylvania at http://www.freerepublic.com/focus/f-news/1898712/posts