See Maniraptora - avian-style semi-lunate carpus allows for greater range of motion in the wrist. Presumably this evolved hand-in-hand (pun intended) with pennaceous wing feathers - having them and being unable to fold up the forelimb in the avian fashion would expose the feathers to damage. This is getting into speculative territory, but if pennate wing feathers evolved primarily for display and/or for purposes of brooding, then there is an imperative to protect them & hence evolve the semilunate carpus.
From: firstname.lastname@example.org <email@example.com> on behalf of Tim Williams <firstname.lastname@example.org>
Sent: Monday, March 27, 2017 2:17 AM
Subject: Re: [dinosaur] Dinosauria reclassification joins Ornithischia and Theropoda in Ornithoscelida
Ruben Safir <email@example.com> wrote:
> in essence there is a cost to the forelimbs not being modified
If by "modified" you mean shortened - then it's been argued that there
is a cost to bipeds (especially cursorial bipeds) in retaining long
forelimbs (e.g., Carrier et al., 2001; Persons & Currie, 2017). This
cost can be mitigated by evolving ways to fold/flex the forelimbs
close(r) to the body.