Since D only claims that glass flows "under its own weight", I think it is actually better supported by the beginning of paragraph 3, which tell you that even solid glass flows, just extremely slowly. If even solid glass flows, it's unlikely that glass at transition temperate does not flow...