Historical masonry structures exhibit pronounced spatial variability in mechanical properties due to material heterogeneity, construction techniques, and long-term deterioration processes. Although the integration of diagnostic data into structural assessment has been widely recommended, the influence of different information acquisition strategies on numerical modelling outcomes remains insufficiently explored. The study hereby presented is embedded within a broader framework in which expert-defined in-situ testing campaigns, designed for a historical masonry component under different budget constraints, constitute the entry point for generating spatially varying mechanical property fields. The resulting property maps, characterised by different levels of uncertainty and spatial coherence, are subsequently propagated into numerical models. A masonry wallet is adopted as the reference structural element, enabling a controlled investigation of the effects of spatial variability while remaining representative of masonry behaviour at the component scale. Numerical simulations are performed using a damaging block-based approach under compression, shear, and flexural loading conditions, considering the multiple spatial variability scenarios previously derived from expert-defined information acquisition strategies. The numerical analyses enable a comparative evaluation of how different diagnostic approaches influence predicted structural behaviour, strength, and damage patterns, relative to a reference benchmark. Rather than focusing on absolute performance, the study highlights trends in numerical sensitivity to spatial data quality, providing insight into the extent to which increased diagnostic accuracy translates into meaningful improvements in structural response predictions. By linking diagnostic strategies to numerical performance indicators, this work contributes to the development of informed and potentially optimised information acquisition strategies for the structural assessment of historical masonry, contributing to more rational decision-making within data-informed Digital Twin frameworks.