Using Spetners definition (his expertise is in information theory) of what
constitutes new information (and he deals in his book with each of the
cases you suggest), none of these meet the criteria of new information.
Either they exploit information that is already in the genome, but
inactive, or they involve a loss of information that decreases specificity.
You can test your ideas against his very careful explanations, if you wish.
Art
http://biology.swau.edu