Deep Eutectic Solvents (DESs) are an emergent class of liquid with a host of useful properties including but not limited to low volatility, cheap components, and a wide chemical design space. These properties have found them a home in numerous applications such as redox flow batteries, biomatter synthesis and processing, separations, and catalysis. Despite their broad appeal, a majority of DES applications are \ extit{ad hoc} and little is known about the fundamental, molecular-level interactions that effect their useful bulk properties. Type III DEss, composed of a hydrogen bond donor (HBD) and a hydrogen bond acceptor (HBA) pair are the most frequently studied with many focused on the prototypical DESs Reline, Ethaline, and Glyceline. There still remains the question of understanding the entire concentration range of a DES, not only the eutectic point, as well as investigating more complex HBDs. Computational modelling offers a way to delve deep into these fundamental interactions, especially when combined in a collaborative effort with experimental experts. Fully atomistic classical molecular dynamics (CMD) in particular is well positioned to simultaneously capture dynamic and structural features to shed light on structural and dynamical heterogeneities present in these novel systems. This is the primary focus of this work, using CMD to study in depth (i) glycerol-choline chloride DES mixtures of varied concentrations to explain DES-unique characteristics and (ii) phenolic-derivative HBDs mixed with choline chloride to derive an understanding of how HBD molecular structure influences DES bulk behavior. The secondary focus is on utilizing quantum mechanical density functional theory methods to understand and predict redox molecule stability in an alcohol-based DES. Fukui functions and dual descriptors use these high-level calculations to explain reactivity differences in two oxoammonium cation molecules that are potential redox molecules in a redox flow battery based on DESs. A further application of condensed Fukui functions and dual descriptors allows for quantitative assessment of stability, allowing for a scalable and potentially automatable analysis of redox molecule candidates.