3) limited, but well-expressed development in a digit can cause distal condensation to cease, in which the more proximal elements take up the role of the distal most elements, usually unguals. This occurs in mammal digits, and why even carnivorans have claws despite thier 2-3-3-3-3 phalangeal pattern.
5) duplication events can cause the identity of a bone to be repeated in series, say for an interruption then later re-expression of the control genes on the growth plate, but only in series, hence the polydactyly of some amphibians, ichthyosaurs, etc.
<And there are better matches in the protorosaurs than anywhere in the archosaurs -- at present. They just haven't been input that often.>
This is based on a biased opinion no one else has yet supported in print since 2000, either the interpretation of pterosaurs or the relationship with "prolacertiforms" or "protorosaurs."