The most primitive known chordates have no hard skeleton at all. Some of
the urochordates (sea squirts, etc.) have tiny mineralized spicules.
Conodont animals have mineralized teeth, but no other hard parts; there is
some change in the formation of the teeth over time (from simpler to more
complex). The earliest known fish, on the other hand, have mineralized
scales but no internal hard skeleton. A cartilaginous internal skeleton
may have preceeded the fully bony one of most vertebrates, as many
primitive fish have mostly cartilaginous skeletons, but it could be
secondary. There are also skeletal transitions between fins and legs,
jawbones and earbones, etc.
The presence of scales and feathers in modern birds is not very strong
evidence for the transition. I mentioned it simply because of Huse's
ill-informed claim. Early fossil birds have teeth, a long bony tail,
claws, growth lines in the bone, a poorly keeled breastbone, and many other
similarities to advanced archosaurian reptiles not present in modern birds,
but they have feathers, flight, and other features of modern birds. This is
better evidence for transition.
There are two reasons why I think common descent is a better explanation
than common design, not by descent. (God was in control of the descent, so
it is designed, despite the denials of atheists and antievolutionists).
The first is the presence of transitional forms, which are not necessary
for de novo design. Secondly, the relative similarity between organisms is
consistent across many unrelated features. I don't see any reason why
design, when not operating through "natural" means, would not combine some
features of one organism and some of another in ways not easily explained
by convergent evolution.