Open peer commentary on the article “Investigating Extended Embodiment Using a Computational Model and Human Experimentation” by Yuki Sato, Hiroyuki Iizuka & Takashi Ikegami. Upshot: First, we comment on a potential weakness highlighted by the use of self-reporting in the human-coupled windmill experiment as described in the target article. Second, we suggest that the authors treat their windmill models as soft-assembled dynamical systems. This would allow them to investigate extended body schemes by looking for 1/f noise in the interface between (...) the agent and the windmills. (shrink)