The rapid development of mobile sensing technologies has enabled the acquisition of three-dimensional vehicle models with consumer devices. These tools are attractive for road accident investigation because they reduce equipment requirements, simplify field deployment and permit the condition of a damaged vehicle to be preserved in a digital form. This study compares three complete mobile workflows applied to the same damaged Toyota Auris: reconstruction from photographs, reconstruction from video frames and mobile LiDAR scanning. All data were acquired with an iPhone 16 Pro Max. The datasets from photographs and video frames were reconstructed in RealityScan, while the mobile LiDAR dataset was acquired and processed in Polycam. The models were then cleaned, oriented and scaled in CloudCompare using the known 2.600 m wheelbase before export as E57 point clouds and integration into PC-Crash. The assessment considered acquisition and processing effort, model density, surface completeness, representation of collision deformation, dimensional verification and practical alignment in accident reconstruction software. The reconstructed point clouds contained 552,983 points for photographs, 589,372 points for video frames and 584,652 points for mobile LiDAR. PC-Crash verification produced wheelbase values of 2.601 m, 2.606 m and 2.601 m, respectively. Reconstruction from photographs provided the most complete surfaces and the clearest local detail. Video frames reduced field effort while retaining satisfactory vehicle geometry, although larger gaps were visible on the roof and around glazed areas. Mobile LiDAR provided a direct and practical workflow but represented thin elements and reflective surfaces with less detail. All three models remained suitable for practical alignment and visual engineering assessment in PC-Crash. An Engineering Workflow Assessment Matrix is proposed to compare the methods through transparent practical criteria rather than relying only on geometric registration metrics.