The small‐island effect: fact or artefact?