Above: Illustration of the megaraptoran Australovenator by T. Tischler, licensed.Well, that all came together...
The history of Megaraptor is long and sordid. It was first described as a gigantic dromaeosaurid (hence the name), on the basis of a large sickle-claw found in Argentina. In fact, I remember a time when rumors circulated that the newly described Unenlagia might turn out to be a juvenile form Ironically, an ACTUAL gigantic unenlagiine dromaeosaurid from almost the same time and place was later found, in the form of Austroraptor. Not to be confused with Austrovenator, which, see below...
Then, a complete Megaraptor forelimb was found that crushed the dreams of every 12 year old Jurassic Park fanboy who had already adopted Megaraptor as their personal mascot: The giant claw came from the hand, not the foot, and it was no raptor at all: but what was it? The first thing that comes to mind is the similar story of Baryonyx ("heavy claw"), named for a similar giant claw that was similarly thought to come from the foot of a dromaeosaur, and similarly later found to come from the hand (have we learned nothing?). Baryonyx was actually a spinosaur, and the large hand claw seemed to be a unique feature of that group. So, was Megaraptor a spinosaur? Opinion was divided for years, with Megaraptor being placed either among the spinosaurs, the carcharodontosaurids*, or in some weird, new family of tetanuran theropods.
The picture got clearer with a few discoveries in just the past year from 'Stralia. First, a large claw that looked suspiciously like the hand claw of Megaraptor, followed by a more complete specimen of a similar animal, named Australovenator. Australovenator looked like a good cantidate for a Megaraptor relative, and studies showed it was in fact an allosaur, somewhere between Allosaurus and carcharodontosaurids.
Well, a new study online today helps clarify the situation a whole lot. Benson, Carrano and Brusatte ran a massive phylogenetic analysis of allosauroids, and recovered a monophyletic clade that happens to include almost every oddball misfit theropod in the book. Called Neovenatoridae, this new clade includes the titular Neovenator (at times thought to be an advanced allosaurid or primitive carcharodontosaurid) and the enigmatic Chilantaisaurus, also previously thought to be a spinosaur. Neovenatorids more advanced than those two are placed in the advanced unranked clade Megaraptora (a nice compliment to the dromaeosaurid clade Microraptoria), which includes Megaraptor and Australovenator as well as perennial phylogenetic stragglers Aeroseon, Fukuiraptor and Orkoraptor. This true identity for Orkoraptor (previously thought to be a coelurosaur) is particularly interesting, since it lived so recently--only 70 million years ago, close to the end of the Mesozoic, showing that derived allosaurs lived right to the end, rather than dieing out in the mid-Cretaceous as previously thought.
All in all, this new paper is pretty major. It reveals a new clade of theropods of a type not really known before, even though none of the constituent species is actually new. What all the megaraptors have in common is a light, gracile build with often hollow bones (seen in the extreme case of Aerosteon), relatively long front and hind limbs (indicating a fast-running lifestyle) and large, hooked claws on the hands for grappling prey. Basically, this advanced line of "coelurosaur mimicking" allosaurs survived in the southern hemisphere while coelurosaurian carnivores were dominating the northern continents, during and after the time of their traditionally carnosaurian relatives carcharodontosaurids.
* I had to stop myself from writing "carcharodontosaurian" instead of "-id" a few times. See, neovenatorids and carcharodontosaurids are sister groups, and the clade containing both has been named Carcharodontosauria. So, technically, Megaraptor is a carcharodontosaur after all, just not a carcharodontosaurid proper.