Why do you think that being a bird somehow eliminates Archy from consideration from being transitional? Aside from the strawman view of transitionals made clear by the moronic 'half and half' description.
You first assume evolution and then see if you can make it fit.
Testing a hypothesis means using that hypothesis to make predictions. If descent with modification is true, then there has to have been species with features of both ancestral forms and later derived forms, some of which which should be found. Note 'half-and-half' is not what is predicted nor is it required. Whether it's considered a bird with many saurian features or a birdlike dinosaur, Archy is clearly an example of what that part of the theory of evolution predicts should be found.