Only branching order is informative in a cladogram, not necessarily
the order taxa are in from top to bottom. Remember, the tree above is
exactly the same as
0--+--Longisquama
| |--Megalancosaurus
| `--Coelurosauravus
`--+--[other diapsids]
`--Icarosaurus
or any other permutation of order you might think of that still retains
the same branching.
Understood.
[...]
It should raise red flags for you when such specialized and highly
derived forms (rib gliders, tail hookers, plume raisers) are nested
(1) so close together without any demonstrated morphological
similarities and (2) without any linking taxa and (3) so close to the
base of a much larger and less well-endowed clade. You're putting the
Styracosaurus before the Psittacosaurus.