3D Hypothetical Reconstruction as a Scientific Process: Integrating 3D Modeling and XR Visualization Within the Critical Digital Model Framework

Recent advances in digital technologies are transforming the production and visualization of 3D models for Augmented Reality (AR), Virtual Reality (VR), and cultural heritage reconstruction. Within this evolving context, ensuring scientifically grounded, transparent, and interpretable reconstruction processes remains essential, particularly for the hypothetical reconstruction of lost or unbuilt architecture. This article discusses the theoretical framework defined by the Critical Digital Model (CDM) and the Scientific Reference Model (SRM), both of which aim to define the 3D model as a scientific product generated through a transparent and falsifiable research process. The proposed approach integrates a structured reconstruction methodology based on source analysis, semantic segmentation, and iterative validation, distinguishing between Raw and Informative Models and applying these principles to selected case studies. Attention is devoted to visualization as an integral component of the reconstruction process. Rather than serving solely as a presentation tool, visualization is examined as a means of analysis, interpretation, and communication of uncertainty. Different visualization strategies—including photorealistic, non-photorealistic, uncertainty-driven, and diplomatic representations—are assessed together with XR visualization modalities, from static images and spherical panoramas to fully interactive VR environments. The results demonstrate how immersive and methodologically grounded visualization approaches can enhance both scholarly investigation and public dissemination, supporting informed choices according to specific research and communication objectives.