Interesting to see that as of writing this, this book has nothing but 1-star and 5-star reviews. I hope that my review can offer a slightly different perspective.
As others have pointed out, Feduccia's primary arguments rest on ridiculous assumptions, completely unsupported reinterpretations of data, and to be frank, basically devolve into petulant whining and vitriol in the latter parts of the book. This has been discussed in detail by reviewer Herman Diaz above, so I won't rehash his points unnecessarily here. Instead I'll focus on the aspects of the book that prevented me from giving it the lowest review score possible.
Early in the book, Feduccia makes some legitimate points that I think should be carefully and unemotionally considered by the scientific community. It absolutely is the case that journals have a bias for what kind of research they accept. Feathered dinosaur-related research is very mainstream right now, and the media and the general public just eat it up. Journals and institutions love the financial support and attention this kind of publication incurs, and so I do not doubt that they are more keen on publishing research that supports the dinosaur-bird connection.
And by the same process, I also don't think it's surprising that journals are less willing to publish research that may result in conclusions that go against the dinosaur-bird dogma: it's less talked about, and less considered, and less interesting to the public. I also think Feduccia is correct in his assessment that the true evolutionary origin of birds is still somewhat shrouded in mystery - in the details, at least. The evidence at this point in time firmly indicates that birds diverged from coelurosaurian dinosaurs, but when? Where? Ground-up or trees down or somewhere in between? From gliders, or from animals who had nothing to do with gliding? Through media representation of research, the general public has adopted the idea that the question of bird evolution and the evolution of flight is completely solved, open-and-shut, when in reality it couldn't be further from the truth. For instance, there is an Triassic trackway from Argentina that, from all appearances, seems to be of bird tracks with fully-reversed halluces. The only follow-up study (Genise et al 2009) done on this ichnotaxon seems to indicate that the tracks involved evidence of takeoff and landing, which further supports the idea that these were definitely birds. While this seems to be (unfortunately) a popular line of evidence among BANDits (and creationists, sadly), there is no mainstream explanation for how these tracks exist other than a vague and unsatisfactory "the age needs to be established with more certainty before we can conclude anything." But rather than do additional studies and research into analyzing this trackway which could be really interesting, mainstream paleontologists seem to have basically forgotten about it.
Along a similar vein, Kenneth Carpenter (1997) has pointed out evidence of Gorgosaurus scale imprints that have been known for at least twenty years, but have never been formally published. Research can of course take many years to publish for a myriad of reasons, but it seems highly likely that had these imprints been of feathers, they'd been published almost immediately. It seems like there is something fundamentally wrong with a system that more readily publishes research that is exciting and interesting because it conforms so smoothly with the dominant paradigm, when conflicting research that challenges some of these established lines of thinking might ultimately result in a more robust and less flawed theory overall.
Feduccia also repeatedly criticizes the practice of cladistics throughout the book, and indeed all of his theories more or less necessitate the invalidation of cladistics, since they completely fly in the face of the current understanding of birds being nestled within maniraptorans. Again, this is a wasted opportunity. There are, I think, some reasonable critiques that can be made of cladistics, since its reliability is completely dependent on a number of subjective interpretations: which characters are worth including, whether each character is truly present or absent, which organisms are included; it also cannot take into account convergence or reversals. But imperfect though it may be, cladistics is ultimately self-correcting, given enough time, enough material, enough organisms and enough revisions, and most importantly, it's the best logical system we currently have for establishing relationships among extinct animals (when used alongside common sense, and when knowing its limitations). Feduccia does not seek to present a consistent alternative to cladistics, nor does he seek to address with any degree of specificity what particular mathematical or anatomical errors are committed in cladistics repeatedly assigning birds to be a nestled group of dinosaurs.
I think Feduccia is a brave guy for bringing up these uncomfortable ideas, a sparse few of which I believe do genuinely deserve a second look. Unfortunately, he also seems to be pretty misguided, likely too incensed by his emotional attachment to his BAND ideals to objectively look at the evidence, which is so much more voluminous than it was thirty years ago or whenever he started writing about this (when it was still a valid alternative hypothesis). Nevertheless, it's really too bad that we don't tend to see this kind of criticism on journal bias and the blind adherence to cladistics outside of "fringe" ideas like BAND. I think these are legitimate issues, and I wish that more "mainstream" scientists would take a look at them, but no one besides the fringers really seem to care. I think that non-BANDit paleontologists have become almost sort of afraid of researching anything that might not fit neatly inside of the current birds-are-dinosaurs thinking, lest others think they're secret BAND supporters, or unwittingly lending them more credibility than they deserve. In reality, the opposite should be the case: every piece of conflicting research, if carefully crafted, should only serve to strengthen a sound theory in the end.
Anyway, my main point is: I don't think BAND is entirely crazy, I just think they're maybe 90% crazy. But they are not crazy on the same level as creationists, and while I certainly understand and sympathize with mainstream paleontologists for how annoying and dishonest they can be, I also think it's too bad that no one seems to really be making an effort to take seriously the valuable things BANDits have to say, while discarding the crap.
