Many species of butterlies and moths eat honeydew, aka aphid poop, so no nectar needed. Also the larvae do the bulk of the eating. Imagos of lepidoptera only need to survive not grow so food requirements are fairly low.
This specimen comes from a communal latrine in the lower Carnian deposits (∼236 Ma) of the Chañares Formation, La Rioja Province, northwestern Argentina. The tiny fossil scales are hollow and ornamented, which is a synapomorphy of Lepidoptera and suggests that they could belong to this group. If this is the case, the Chañares scales would partially fill the temporal mismatch between phylogenomic date and the fossil evidence of butterflies and moths because they preceded the previously oldest lepidopteran record by c. 35 million years. Moreover, the scales have a combination of features present in early diverging glossatan lepidopterans. The inclusion of the temporal data provided by the Chañares scales into an updated temporal calibration of lepidopteran phylogeny shows that the proboscis, a key evolutionary novelty for the group (Glossata), evolved between c. 260–244 Ma. Thus, the proboscis-bearing lepidopterans would be part of the repertory of new plants and animals that diversified during the aftermath of the EPME.
Realistically speaking, would they even be able to support a body that large on those types of wings? Or would they have to evolve some form of rigid structure like a bird or bats wings?