Recent detail-preserving shape deformation techniques are either based on a combination of multiresolution decomposition and variational bending energy minimization, or they manipulate differential coordinates and solve a Poisson system to obtain the deformed surface. We establish an explicit connection between the two approaches and discuss their inherent limitations, such as local selfintersections for the former and translation insensitivity of the latter. Based on these new insights we combine both methods into a novel shape editing technique that does not suffer from previous limitations, while retaining editing flexibility and efficiency.