So I did a little googling. The authors of the paper think those proboscises evolved originally to drink water, and some butterflies and moths then started using them partly to get nectar from horsetails. There's also an offhand reference (as to something that people in the field already know) to "increased herbivory" of insects, with specifics about leaf-eating.

Meanwhile, Scientific American also has people saying "maybe not" about significant ancient diversification of Lepidoptera, but it's not clear whether they mean "no, you're trying to get too much from a few fossil scales" or "yes, these three lineages go back to the end of the Triassic, but that doesn't mean there were lots of moths and butterflies flying around the early Jurassic landscape."