The Origins of 'Avian' Characteristics

It's very easy to overestimate or underestimate when certain characteristics present in modern birds evolved when depicting stem birds (any extinct taxon closer to birds than to any other extant taxa). Science marches on aside, it is not uncommon to see enantiornithines shown with tail fans and deinonychosaurs shown without pennibrachiae, for instance. Here I've drawn a phylogenetic tree of Archosauria showing where ancestral characteristics of modern birds (with most traits basal to Archosauria as a whole also being present in modern crocodilians, of course) evolved. Many of these are internal features though and aren't readily visible in life. I've mostly chosen to depict basal members from each group pictured to highlight their similarities. Heavily referenced from the cladograms in Dr. Thomas Holtz's dinosaur course syllabus.

Characteristics marked with a question mark are those where it's not clear whether they were first present at that point in the tree. In many cases said traits may have already evolved long before the point marked. (Feathers are the obvious example*; in recent years the idea that feathers may be a characteristic basal to all dinosaurs or even beyond has been gaining support. My personal suspicions at this point are that some sort of feather homologue was indeed present at the base of Dinosauria at least, and was independently developed into an extensive insulating coat by several lineages.)

I've only labeled characteristics where they are actually homologous with the condition in modern birds. Therizinosaurs for example have a first toe that touches the ground like modern birds, but this is almost certainly convergent, to say nothing of the numerous lineages that have evolved rhamphotheca independently.

*The reason I didn't extend feathers further down the tree in the latest update is because the phylogenetic position of Sciurumimus is not all that secure. Okay, and because I didn't have a megalosauroid branch in the initial drawing and didn't want to add it just for the sake of one (equivocal) characteristic.