Conditioning is the primary method for belief revision in data fusion systems employing probabilistic inferencing. However, big-data environments, where soft (i.e., human or human-based) sources are commonly utilized in addition to hard (i.e., physics-based sensors, pose several challenges to traditional conditioning tasks primarily due to the numerous data/source imperfections that are characteristic of such data. The objective of this paper is to investigate the most natural extension of Bayes conditioning based evidence updates in the presence of such large-scale data uncertainties and source/sensor imperfections. By viewing the evidence updating process as a thought experiment, we devise an elegant strategy for robust evidence updating in the presence of extreme uncertainties characteristic of big-data environments. In particular, we look at the formulation of a belief theoretic evidence updating mechanism that is derived as a natural extension of Bayes conditional approach when the incoming evidence takes the form of a general belief function. Proposed method generalizes the belief theoretic Fagin-Halpern conditional notion, and provides a novel evidence updating strategy that is derived as a natural extension of Bayes conditional applied in a highly uncertain and complex fusion scenario that is characteristic of big-data environments. The presented extension differs fundamentally from the previously published work on Conditional Update Equation (CUE) as well as authors own work. An overview of this development is provided via illustrative examples. Furthermore, insights into parameter selection under various fusion contexts are also provided.