Which... isn't really an objection. After all, there is generally no evidence that the taxon concerned "didn't need feathers", even though there may be evidence from phylogenetic inference or even direct fossil evidence that it did have them. Besides, many animals (including many dinosaurs) are known to have some crazy structures for which their function isn't obvious, but that isn't really evidence those animals didn't have them. (Who can actually say, for example, exactly what stegosaur plates or Psittacosaurus bristles were for?) This ends up as more an argument from personal incredulity (i.e.: "I can't think of how taxon A would've used its feathers, so it didn't have any!").
Nevertheless, just as with stegosaur plates and Psittacosaurus bristles, it's still good to try and seek out potential functions for feathers in non-avian maniraptors. Fortunately, the function of feathers is somewhat easier to infer than stegosaur plates or Psittacosaurus bristles, as there are still feathered maniraptors alive today.
All living maniraptors are either flighted or have flighted ancestors, and a primary function of feathers in neornithines is to help them fly. However, no non-avian maniraptor was quite as adapted to flight as modern neornithines, even though some clades have been suggested to have been secondarily flightless and a number of small deinonychosaurs could've had some limited aerial capability. The presence of plumaceous feathers with no aerodynamic qualities in non-maniraptor coelurosaurs with no flight adaptations shows that feathers arose before flight in dinosaurs began to evolve. (That's right, plumaceous feathers were probably present in basal coelurosaurs. Might talk a bit about that later in the series.) As feathers are found in flightless coelurosaurs (including modern ones), they must also be beneficial in other ways. Feathers are exaptations; they were later adapted for flight, but that wasn't their original function.
Another major function of feathers in modern maniraptors (both flighted and flightless) is insulation, and feathers are important in both keeping maniraptors warm in the cold and keeping them cool in the heat. Plumaceous feathers found in basal coelurosaurs (as well as in basal maniraptors such as therizinosaurs and alvarezsauroids) likely had this function.
Many non-avian maniraptors didn't just have plumaceous feathers, however. Oviraptorosaurs and deinonychosaurs are both known have had pennaceous feathers. Pennaceous feathers are also useful for insulation, and asymmetrical pennaceous feathers (which are known in the dromaeosaurid Microraptor) are good for gliding and flying with. In addition, pennaceous feathers are also more useful than plumaceous feathers in visual communication, due to their wider display area and (possibly) greater range of colors that they can show. Some feather structures known in non-avian maniraptors, such as tail fan and the short primary feathers forming little "hand flags" in Caudipteryx zoui and the feather crests in Microraptor and Anchiornis, were likely for display purposes. (In the case of Anchiornis, the feather crest is known to have been orange, in contrast to the dark gray and black of the rest of the animal, which supports its use in visual display. Similarly, the wing feathers of Caudipteryx show a banded color pattern. Furthermore, the development of wing feathers in Similicaudipteryx signify that the wing feathers were more important in older individuals than in younger ones.)
|Male Pavo cristatus in display using pennaceous tail feathers photographed by PRA, from Wikipedia.|
The greater area of pennaceous feathers likely served another purpose, too. Several oviraptorid and troodont specimens have been preserved brooding on top of their nests with the forelimbs extended over the eggs. The wing feathers that were likely present in these taxa would have helped cover the eggs while the dinosaurs were brooding.
|Fossil of Citipati osmolskae brooding on nest photographed by Matt Martyniuk, from Wikipedia.|
Some flightless birds have yet another use for their wings. Rheas and ostriches flick their wings out to the sides while running to help them make quick turns. Cursorial flightless non-avian maniraptors may have done the same.
An interesting idea that's been proposed in the last decade is Wing-Assisted Incline Running (WAIR). Many birds, particularly those that are primarily terrestrial or too young to fly, can escape from predators by flapping their wings to help them run up inclined surfaces such as tree trunks. This behavior was first observed in chukar partidges and has been documented among many neornithine groups. It's been suggested that small non-avian maniraptors with well-developed wing feathers may have also used WAIR and that it contributed to the evolution of flight. Other researchers, however, argue that the restricted forelimb movement in non-avian maniraptors wouldn't have allowed them to perform WAIR. (Non-avian maniraptors couldn't lift the wings above their backs. Indeed, not even confuciusornithids could!)