Are Large Brainwave Foundation Models Capable Yet? Insights from Fine-tuning