It would require two more reversals -- in *Sapeornis* and in Confuciusornithidae. Four reversals instead of two or maybe three acquisitions.
The condition is unknown in _Sapeornis_ and confuciusornithids because the
suturing/fusion obliterates the nature of the articulation. Thus, they
could have evolved from birds with the enantiornithean-style peg-and-socket
articulation, and the immobile joint could be secondary. Under this
scenario, the condition in _Sapeornis_ and confuciusornithids is not a
reversal. It would be (and has been) coded as '?'.