Good to see Gauthier back in the dinosaur business!
It shows that Compsognathus longipes had two fully formed functional digits as well as a reduced, perhaps even non-functional, third digit.
Interesting.
That conclusion runs counter to the usual interpretation that Compsognathus longipes had only two phalanges, rather than the expected complement of three, in digit II.
Gishlick & Gaulthier's Compsognathidae includes only _Compsognathus_ and _Sinosauropteryx_. The paper only looks at the morphology of the hands, and doesn't include a full cladistic analysis.