Small-format multi-head camera systems are increasingly deployed for UAV-based 3D mapping; however, systematic evaluations that explicitly link acquisition geometry to network robustness and façade reconstruction completeness remain scarce. This study establishes a geometry-driven, multi-dataset evaluation framework to assess such systems in topographic mapping. Six UAV datasets with varying flying altitudes, overlaps, platforms, cameras, and urbanization are analyzed using a rigorous bundle block adjustment workflow. Internal network strength is quantified using image connectivity, image ray intersection angles, and tie point multiplicity, while object-space accuracy is assessed using ground control point residuals. Reconstruction completeness is evaluated using façade-specific point density metrics, enabling a controlled comparison between nadir-only and multi-head (nadir + oblique) configurations. The results show that oblique imagery strengthens multi-directional image network geometry and substantially improves façade reconstruction. This study establishes empirical relationships between acquisition design, network stability, and object-space completeness.