Revisiting Automatic Speech Recognition Systems for Augmentation Capacity