These features are eminently evolvable, as it were, quite within the grasp of normal means by which heritable variation is generated.
Not in a way that is coherent, I describe the problem in slides 99-110 of part 1. Briefly, homologous motif duplication or whatever duplication will result in changes in binding specificity, worse, random mutation will cause the zinc finger to jump to different locations on the genome, essentially binding to random locations for DNA-binding zinc fingers. We wouldn’t want complexes like that pictured in slide 95 docking to random locations on the DNA and recruiting who knows what kind of chromatin modyfying complexes and making who knows what kind of chromatin modifications!
EDIT: Slide 112 shows my conclusion that the “phylogeny” is imaginary as a matter of principle.