> "birds" are essentially defined as the set of theropod species that survived the K-T extinction

The problem with that is that this bird would not be considered a bird, since it didn't survive the K-T extinction. It is considered a bird because it is bird-like enough. (and the concept of "bird" is a man-made category...and there would be a lot of blurriness if you look at species that lived at the same time as (other) dinosaurs.

You're right, insofar as "bird" is not jargon, or in 1:1 correspondence with jargon, in any scientific discipline. There's just theropods, some of which happened to survive, where people call the descendant species of those particular theropod species birds. There's no rule specifying whether those theropod species themselves are birds, or whether species that are 99% like them but happened to die off due to pure bad luck are birds, etc.—because we don't need such a rule for anything.

If we did, though—if "bird" was jargon under some branch of science, the way that "fruit" is jargon in culinary science†—then I would guess that the criterion for it would be something like "a theropod with adaptations that would make it likely to survive the K-T extinction, all else being equal", making both the ancestor theropods to current birds qualify, as well as the bird in the article. But, well, nobody needs that definition for anything, so it can't really be argued what the definition "really is."

† https://en.wikipedia.org/wiki/List_of_culinary_fruits

"Birds" is a taxon (Aves is the scientific name for it). In modern biology, all taxa are defined to be "monophyletic lineages" or "clades", which means that in principle they comprise one ancestral population and all descendants. No taxon would be defined nowadays as you define it since your definition does not necessarily identify a clade. A valid monophyletic taxon might be coincident with a set of organisms identified by a definition such as yours, however.

Edit: The Wikipedia article has a section on the definition of the relevant taxa: https://en.wikipedia.org/wiki/Bird#Definition

Although, interestingly the article describes the find as belonging to an ancient group of toothed birds, so perhaps the Wikipedia description needs a tweak?

All toothed birds are extinct. Some birds today have serrations on their beaks that might look like teeth, but they're not really teeth in an anatomical sense.

>"birds" are essentially defined as the set of theropod species that survived the K-T extinction.

This is not a definition of birds that is scientific or even colloquial. Nobody defines birds this way.

Also, we only consider the extinction to be "over" (the point at which we can declare who is and is not a survivor) when the only theropods left were birds, so this definition is circular. This is the problem with trying to derive necessary and sufficient conditions for things like "birdness", the concept was constructed by pattern matching, not analytically, so rarely will an analytic definition capture everything that the common sense understanding encompasses.

All of taxonomy is pattern matching unless you have something like kennel clubs or Mormons tracking individual genealogy​

We're not talking about the same "pattern matching" because even those examples are pattern matching. You have breeds of dogs and a dog is either one breed or another or it's a hybrid/mutt. But at what point in the process of creating a new breed do you switch from handling hybrids of existing breeds to handling purebreds of your new breed? No single generation seems like it could be enough to make the difference, since the puppy is just as similar to its parents as the puppies of any other generation, but on the scale of many generations, a new breed clearly emerges. It's analogous to the paradox of the heap [1].

Derek Parfit's Reasons and Persons engages with this problem as it applies to personal identity with his "spectrum cases".

[1] https://en.m.wikipedia.org/wiki/Sorites_paradox

edit: too late to delete. I misunderstood the OP.

Appologies.

That was kind of the point of my definition, actually—it's easier to predict what type of thing is a bird if you think of "bird" as "the clade rooted on all the theropods that fit through a particular evolutionary sieve" rather than a category with certain characteristics. Not all modern birds have those characteristics. But those characteristics are what allowed the prototypal "birds" to survive where other dinosaurs did not. Their descendants went on to adapt in various other ways, but we still consider them all "birds" for reasons of genetic lineage back to those original birds.

Sorry, I misunderstood you and thought your "flight" point was meant to be valid for current species as well as the early incarnations. Appologies.

Yes, they evolved later (generally under non-mammalian island biogeographies which produced evolutionary pressures which meant that being large and foraging nocturnally had benefit over retaining flight).

This happens all the time -- imagine the evolutionary pressures on proto-cetaceans. Now imagine writing a comment discounting the benefits of land-dwelling because some paths returned to the sea.

Indeed. However where I erred was misunderstanding the GP. I thought his "flight" description was intended to be relevant to present birds as well.

Flightless birds are birds that lost their capability to fly.