Other Velociraptor forearms are quite smooth—no knobs. Assuming all members of this group had feathers that required such strong anchoring, shouldn't at least the well-preserved ones show quill knobs?
Do you have a source for this? (I mean a photograph or description of a specific fossil, not just what ICR says.) There are not many
Velociraptor fossils where the arm bones are well-enough preserved for it to be possible to tell whether it had quill knobs or not. Being able to make this sort of judgement about quill knobs doesn't just require that the arm bones be preserved; it also requires that the bones haven't been significantly distorted or worn after the animal died.
In the specimen where the quill knobs were identified, identifying them required an electron microscope. As far as I know, no one has done that test on any of the few other
Velociraptor fossils that are well-enough preserved to possibly make that judgement about them.
These next several points are related to one another, so I think it's best to answer them all at once:
The bumps identified as quill knobs on all these dinosaur bones have peculiar differences from the real ones we see today on certain bird bones.
Some are much smaller in proportion.
Some have different or less regular spacing.
Other dinosaur bone knobs run along the outside of their arms, instead of along the back of the bone as in modern birds.
Real quill knobs occur in birds with feathers needing very strong attachments because of the rigors of powered flight. But Dromaeosaurid arms were too small to serve as wings—and in the case of Dakotaraptor the arms were much too small. Why would these creatures need such strongly anchored feathers if they couldn't even use them to fly?
Dromaeosaurids are not direct ancestors of birds, so it's expecting too much if you expect their and birds' quill knobs to be identical in structure. One obvious difference is that with a few possible exceptions (such as
Changyuraptor and
Microraptor), dromaeosaurids weren't able to fly. Birds need especially strong and robust quill knobs, because when they fly their wings have to support their entire body weight, but dromaeosaurids didn't have that requirement.
The
Dakotaraptor paper mentions a few possible ways that it might have used its wings, and there's one in particular that would have required somewhat sturdy anchor points for its feathers (although not as sturdy as would have been required for flight.) Modern birds of prey use their wings for one other function besides flight: they also flap their wings to stay balanced on the ground while holding onto struggling prey animals with their feet.
This paper describes how that the feet of dromaeosaurids share several of the same traits that enable modern birds of prey to do this. If dromaeosaurids (including
Dakotaraptor) used their feet in this way, flapping their wings to stay balanced would have been useful to them for the same reason that it's useful to modern raptors.
Also, tiny bone bumps have more uses than just anchoring feathers. They sometimes mark attachment points for connective tissue. Evolutionary paleontologist Darren Naish dared to express skepticism over quill knobs in a Concavenator from Spain. He wrote in a 2010 blog post, "Animals sometimes have weird, irregularly spaced tubercles arranged in lines on various of their bones, typically located on intermuscular lines (they presumably represent partially ossified attachment sites for tendinous sheets or similar structures): I've seen them on mammal bones and on a theropod tibia."
http://scienceblogs.com/tetrapodzoology/2010/09/09/concavenator-incredible-allosauroid/
Why must tiny bumps on dinosaur bones signify feather attachments if they don't on mammal bones?
There have also been studies of these 'quill-knobs' through high powered electron microscopes and they found that these 'quill-knobs' are found in other extinct reptiles, like Leiocephalus Eremitus, and that they don't actually show feathers but evidence of a naturally occurring phenomena formed in Rigormortis.
http://www.icr.org/article/9024
One of the problems with articles like that one from ICR is that they don't include photographs of the muscle attachment points that they say can be mistaken for quill knobs. If they had included a photograph, I think you'd be able to see that the two aren't all that difficult to tell apart.
Darren Naish's article is more useful, because that one does include photographs. Note that on
Concavenator the possible quill knobs are much more closely spaced at one end of the bone than at the other, and they're also on what would have been the side of the animal's arm instead of the back of its arm. These are valid reason to doubt that what
Concavenator has are actual quill knobs. But Naish specifically points out that in these respects
Concavenator is different from dromeosaurids, whereas the location and spacing of quill knobs in dromeosaurids is similar to what it is in birds. These are the sorts of traits one can look at to determine whether an animal has "true" quill knobs or not.