Dance plays a crucial role in human well-being and expression. To learn dance, transferring motor knowledge across humans is relevant. Several technologies have been proposed to support such knowledge transfer from teacher to student. However, most of such systems applied a pragmatic approach focused on the feedback and the quality of the feedback system and not necessarily on the human mechanisms behind the dance learning process. In contrast, we inquire about the teacher-to-student motor knowledge transfer from the neural perspective to design motor learning wearable systems. We conducted interviews with dance students and teachers using vignettes based on motor learning theory as a discussion base. We derived insights about dance learning and identified a series of requirements for motor skill transfer-focused wearable devices. Based on our results, we present a prototype that reflects the minimum functional setup for effectively supporting motor learning.