A majority of drugs marketed in amorphous formulations have a good glass-forming ability, while compounds less stable in the amorphous state still pose a formulation challenge. This work explores ternary solid dispersions of two model drugs with a polymer (i.e., hydroxypropyl cellulose) and a coformer as stabilizing excipients. The aim was to introduce a computational approach by preselecting additives using solubility parameter intervals (i.e., overlap range of solubility parameter, ORSP) followed by more advanced COSMO-RS theory modeling. Thus, a mapping of calculated mixing enthalpy and melting points is proposed for in silico evaluation prior to hot melt extrusion. Following experimental testing of process feasibility, the selected formulations were tested for their physical stability using conventional bulk analytics and by confocal laser scanning and atomic force microscopy imaging. In line with the in silico screening, dl-malic and l-tartaric acid (20%, w/w) in HPC formulations showed no signs of early drug crystallization after 3 months. However, l-tartaric acid formulations displayed few crystals on the surface, which was likely a humidity-induced surface phenomenon. Although more research is needed, the conclusion is that the proposed computational small-scale extrusion approach of ternary solid dispersion has great potential in the formulation development of challenging drugs.