Recent advances in geostationary imaging have enabled the derivation of high spatiotemporal-resolution cloud-motion winds for the study of mesoscale unsteady flows. Due to the general absence of ground truth, the quality assessment of satellite winds is challenging. In the current limited practice, straightforward plausibility checks on the smoothness of the retrieved wind field or tests on aggregated trends such as the mean velocity components are applied for quality control. In this paper, we demonstrate additional diagnostic tools based on feature extraction from the retrieved velocity field. Lagrangian Coherent Structures (LCS), such as vortices and transport barriers, guide and constrain the emergence of cloud patterns. Evaluating the alignment of the extracted LCS with the observed cloud patterns can potentially serve as a test of the retrieved wind field to adequately explain the time-dependent dynamics. We discuss the suitability and expressiveness of direct, geometry-based, texture-based, and feature-based flow visualization methods for the quality assessment of high spatiotemporal-resolution winds through the real-world example of an atmospheric Kármán vortex street and its laboratory archetype, the 2D cylinder flow.