Joined: Jan. 2006
It seems that their definition of information actually refers to the phenotype and specifically the "function", but not fitness.
For instance, antibiotic resistance would be a loss of information if the altered protein is rendered less effective in the absence of antibiotic. I doesn't matter for them that the bacterium is much fitter otherwise. Gene duplications often generate proteins with new functions. However, this is rarely observed at the intraspecific level, and the IDiots might claim that the duplication was designed.