Mildred Dresselhaus's invited talk focused on thermoelectrics and carbon materials (CNTs, graphene). It's always a pleasure to hear Prof. Dresselhaus talk. She has an endearing humility that makes her work and accomplishments even more outstanding. Regarding thermoelectrics, she pointed out how nanostructuring adds another parameter (dimension) that affects the key properties which make up the figure of merit. Her historical perspective on studying carbon was entertaining. It pointed out how new ideas like the metal/semiconductor designation of CNTs based on chirality aren't always embraced at first, but they may later be widely accepted. After all, what CNT lecture nowadays doesn't include the zigzag vs. armchair image?
There wasn't a clear answer to the question in the talk's title. When you probe the length scales that influence material properties, you inevitably probe the nanoscale... so maybe nano is always the way because all materials have nanoscale characteristics that we might like to manipulate.