Will It Blend? Mixing Training Paradigms & Prompting for Argument Quality Prediction