Just saw this new study of Saltopus elginensis which includes a new phylogeny. Now, granted—new phylogenies come out all the time. Many of them to not stand the test of time, and fall. This may be one such too, but until then, it raises an interesting spectre that we've maybe misinterpreted the spread of dinosaurs altogether.
What if the classic division of Saurischia and Ornithischea isn't really correct? This new study suggests that Therapoda (minus Herrerasauridae) is actually to be allied with Ornithischia instead of Saurischia, and he creates a new clade Ornithoscelida. Which is actually the revival of a 125 or so year old proposal by none other than Thomas Huxley.
If true, this would clear up a lot of confusion about the Herrerasaurs. Nesbitt's huge cladogram, which was kind of the gold standard when it was done in 2011 or so (and which had the other surprising discovery of recovering phytosaurs outside of Archosauria) had them as basal therapods, but other cladograms had recovered them as basal saurischians, or even as basal dinosaurs that predated the saurischian-ornithischian split. At least phylogenetically, if not necessary temporally (actually the age of herrerasaur bearing rocks is a problem too.) There have even been more radical suggestions; such as that they were sauropodomorphs, or even dinosauriformes outside of Dinosauria proper. But in this scenario, they are saurischians that are not closely related to therapods at all, because they're on the other side of the Great Dividing Line in dinosaur systematics from them.
Which is also curious, because it would mean that meat-eaters actually popped up on both sides of the line too—which was not the case previously—and which suggests that the original dinosaurs might have been omnivorous. It's also curious that this is in many ways the complete opposite proposal as Phytodinosauria of Bob Bakker fame, which proposed that sauropodomorphs and ornithischians were the more closely related ones, and therapods were distinct. To me, it seems clear that the exact nature of the relationships between these animals is difficult to determine because of their gross similarity, which has led many a researcher astray—regardless of how it all turns out in the end.
As an aside, the data that really changed the analysis were greatly improved descriptions of two primitive and early ornithischians: Heterodontosaurus and Lesothosaurus. Both are believed to possibly have been omnivorous early ornithischians.