Additional Findings on the Multi-Modal Demands of “Voice-Command” Interfaces