I'm BrassButton's paleontologist friend.
Here are a few of my favorite dinosaurs:
1) Sue. I saw her come out of the rock, and I've been in love ever sense.
2) Therizinosauroidea. Imagine a T. rex with scythes as claws. What hope have mammals but for the mercy of the reaper-dino? We don't know why they had them, either. Speculation abounds, including speculation that they used their claws to eat termites (I find that argument HIGHLY dubious--these things weren't small, and it'd take a few trillion termites to make a good meal for them).
3) Heterodontosauridae. I was introduced to these via a paper Paul Sereno did earlier this year, and that was one of the times I've just stared, unable to assimilate the data. Honestly, if an Elder God had stopped me in the street and asked for directions I'd have been less surprised. These things had canines and teeth that look for all the world like horse teeth. I still can't figure out why (and neither can anyone else). Right now, the best guess is that they used them sort of like pigs, to root around and rip into things.
Fun trivia: T. rexes had the strongest bite pressure of anything ever measured. They also had teeth that were banana-shaped and which had serrations that captured muscle fibers and other animal bits. That means that not only did they have a bit that could turn your bones to shrapnel, they'd also infect you with all sorts of fun diseases when they did so. The conclusion for hunting behavior is obvious: they'd ambush prey, bite hard and fast, then back up and wait.
The chicken thing is Jack Horner's pet project these days. And it's not chickens, it's emus. Basically, he's playing with the genetic code to re-activate genes that were shut down (for various reasons), giving emus long tails, claws, scales, teeth, and all the other accoutrements of Therapoda. As an aside, I've got an uncle that used to raise emus. He has thus far refused to allow me to put one of these on his farm.
The whole Ceratopsian line is currently one giant snarl, and it's actually cast some doubt as to dinosaur taxonomy (the finer points, at least). Paleontologists recently realized that what they'd been calling different species were actually different ontological stages--essentially, they were saying puppies and dogs were two different species! They figured this out through analysis of bone growth patterns, something that's not hard to do physically (you make a thin section of the bone, then look at it under a microscope--standard stuff), but which is nearly impossible from an administrative perspective (NO ONE wants to let ANYONE cut up dinosaur bones). So we're sort of waiting to see the fallout of all that.
I recently heard about evidence for Triceratops heads being ripped off their bodies. Apparently T. rexes would grab their neck frill and pop the head off like an old-school beer tab. Haven't gotten ahold of the original paper yet, but I can only imagine that this only occurred after a sufficient period of deterioration--it's VERY hard to rip something's head off!
I actually spent a few years studying the K/Pg mass extinction. It's one of two mass die-off style mass extinctions, with elevated extinction rates but steady or reduced origination rates (the other option is for both to be elevated, which I call mass turn-over style mass extinctions). Studying that extinction ushered in a new sedimentological paradigm: neocatastraphism, which argues that low-amplitude, high-frequency events are typically overwritten in the geological record by high-amplitude, low-frequency events. A great example is a beach: one storm can re-work a few tens of feet of sediment into a single tempestite, basically wiping out a few hundred years of sedimentary features. (I mean no offense here--I know that the devastation from Sandy is ongoing. This is the classic example, nothing more.) And contrary to what you'll hear, there's no actual evidence that dinosaurs were in decline leading up to the impact event. Basically, the reason it looks like there is is because 1) most paleontologists don't pay adequate attention to highly degraded bone or stratigraphic position, and 2) when you have events on a line with randomly-sized gaps, you can stop the line anywhere and it will always give you the appearance of a gradual decline. It's called the Signor-Lipps Effect. Peter Ward did some research on various organisms as they approached the K/Pg boundary back in the '90s, and found that pretty much universally they were doing fine up until the rock hit us. Then everything went to Hell.
One of my most prized possessions is "The Dinosauria", second edition. It's a book that's about five inches thick, and basically gives the state of dinosaur research up until 2004. And I mean "dinosaur research"--EVERYTHING. It devotes page after page to the most god-awful boring anatomical descriptions of bones you've ever seen, everything from the skulls to the wrist and ankle bones. $50 at Barns & Noble. It's a prized possession because it's taken me 23 years to get to the point where I can understand it enough to justify owning it. There have been a lot of nights when I've just stayed up reading those god-awful anatomical descriptions of bones. I highly, HIGHLY recommend this book to anyone who's interested in dinosaurs. It's become my go-to resource (with, of course, marginalia discussing updates since it's nearly a decade old now

).