Simulating a hierarchical, task-based, state-feedback model of speech motor control: insights and issues