The second type is represented by groups of six or seven downwards projecting up to 1.5 centimetres long filaments, together originating from a base plate. These are present on the upper arm and thigh. They resemble the type 3 feathers of theropods. The base plates are ordered in a hexagonal pattern but do not touch each other.However the supplementary material tells a different story:
We considered attempting to describe the feather morphotypes in Kulindadromeus using the nomenclature of Prum et al. (52, 53) or of Xu et al. (21, 22). However, except for our monofilaments (which correspond well to Type 1 in Xu et al.), we could not assign with confidence the other two feather morphotypes in Kulindadromeus to categories described by Prum et al. or Xu et al. Further, fundamental discrepancies between these two previously published nomenclature systems remain to be resolved. Thus we felt that until new fossil material and a synthesis of existing nomenclature systems are available, interpretations of direct homologies between complex feather-types in Kulindadromeus and in Prum et al. or Xu et al. would be premature.A claim of anything beyond "type 1" filaments (that are actually just bristles) goes beyond the evidence, as confirmed by the supplementary material.
This quote expresses exactly what I am saying:
So I find it quite strange and disheartening that Godefroit et al.—despite being fairly objective in their supplementary material—go completely gung-ho in calling these structures feathers.