Reversions

The dino to bird theory requires "remarkable" reversals. 

ANKLE:

http://www.nature.com/ncomms/2015/151113/ncomms9902/full/ncomms9902.html (2015)

The anklebone (astragalus) of dinosaurs presents a characteristic upward projection, the ‘ascending process’ (ASC). The ASC is present in modern birds, but develops a separate ossification centre, and projects from the calcaneum in most species. These differences have been argued to make it non-comparable to dinosaurs. We studied ASC development in six different orders of birds using traditional techniques and spin–disc microscopy for whole-mount immunofluorescence. Unexpectedly, we found the ASC derives from the embryonic intermedium, an ancient element of the tetrapod ankle. In some birds it comes in contact with the astragalus, and, in others, with the calcaneum. The fact that the intermedium fails to fuse early with the tibiale and develops an ossification centre is unlike any other amniotes, yet resembles basal, amphibian-grade tetrapods. The ASC originated in early dinosaurs along changes to upright posture and locomotion, revealing an intriguing combination of functional innovation and reversion in its evolution.

Also see here:

More remarkably, however, this finding reveals an unexpected evolutionary transformation in birds. In embryos of the landegg-laying animals, the amniotes (which include crocodilians, lizards, turtles, and mammals, who secondarily evolved live birth) the intermedium fuses to the anklebone shortly after it forms, disappearing as a separate element. This does not occur in the bird ankle, which develops more like their very distant relatives that still lay their eggs in water, the amphibians. Since birds clearly belong within landegg-laying animals, their ankles have somehow resurrected a long-lost developmental pathway, still retained in the amphibians of today -- a surprising case of evolutionary reversal.

WRIST:
http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1001957

We confirm the proximal–posterior bone is a pisiform in terms of embryonic position and its development as a sesamoid associated to a tendon. However, the pisiform is absent in bird-like dinosaurs, which are known from several articulated specimens. The combined data provide compelling evidence of a remarkable evolutionary reversal: A large, ossified pisiform re-evolved in the lineage leading to birds, after a period in which it was either absent, nonossified, or very small, consistently escaping fossil preservation.

FINGERS:
http://www.nature.com/articles/nature08124.epdf?referrer_access_token=1LIOYM249T2ALXmHhUVXQtRgN0jAjWel9jnR3ZoTv0NAxxXDTxDgb7tt7vNCs5i7CDx_p1E8pIL0dPMGIw0CIZ1LRnUZIDT1a3FIDY_UW4FRwpODRDVwWg-KbK448VK63yIXiGAa_H8fA42yVK8TsNhr_ASjWKKTbM-PJCMVzpKKElR4FEstewHl9DZGaHr9&tracking_referrer=www.nature.com

Based on this study, the most parsimonious alignment is for the four digits of ceratosaurs to be I-II-III-IV and the three (and sometimes four) digits of all Tetanurae to be II-III-IV(V). Accepting such a topological shift at the base of Tetanura requires that the positional homology of the three digits of tetanurans is II-III-IV(-V), as suggested by Wagner and Gauthier34. Because the four digits of ceratosaurs are therefore most parsimoniously interpreted as I-II-III-IV, the small lateral metacarpal ossification of Guanlong35, Sinraptor36, and Coelurus represents the re-ossification of metacarpal V after it is lost at the base of Ceratosauria. The poor phylogenetic resolution for basal tetanurans in our study precludes us from hypothesizing whether this re-ossification event occurred once or more than once in the evolution of Theropoda. Likewise, the fourth metacarpal, which is reduced in primitive theropods and bears an unknown number of phalanges in Ceratosauria, re-acquires at least three phalanges in Tetanurans.

This implies the reduction of digit I before the divergence of the Ceratosauria and the
Tetanurae, the appearance of some polleciform features in digit II and the acquisition of a novel phalangeal formula (X-2-3-4-X) early in tetanuran evolution. Both modifications are partially indicated by the manual morphologies of ceratosaurs and more basal theropods. Also, they are indirectly supported by observations in living animals that a digit will display features normally associated with the neighbouring medial digit if the latter fails to chondrify in early development21, that phalangeal counts can vary even within species29, 42 and that secondarily cartilaginous elements can regain their ability to ossify43.

If BDR [Bilateral Digit Reduction] applies to the more inclusive Averostra, as the II-III-IV hypothesis suggests, early stages of tetanuran evolution must have involved loss of the already highly reduced metacarpal I, reduction in the length of metacarpal II, and the reappearance of additional phalanges on metacarpal IV. Both the I-II-III and II-III-IV hypotheses can claim a degree of support from morphological data, but the II-III-IV hypothesis is more parsimonious when developmental data from extant birds are considered.

There is no actual evidence for these reversals. The dino to bird theorists need to imagine they happened so the dino to bird theory does not collapse.