Polymer solar cells (PSCs) are lead-free and can now be processed from relatively benign solvents with high efficiency of >14%. For many years, progress has relied mostly on intuition and trial-and-error approaches. For example, it was conceptually well known that fundamental molecular interactions must control the achievable PSC morphology, quantifying these interactions and predicting important transient and final device morphology parameter such as the composition of the mixed domains and their relation to performance remained unattainable for a long time. This presentation will review the history of this winding road, the lost tracks, the parallel paths, and some of the blind alleys of the field to understand “miscibility” and the phase behavior of PSC blends. Recent progress indicates that we might converge on a coherent understanding of structure-function relationships and that computer simulations might help to screen materials combinations for their suitability of high performing system. In some cases, the amorphous-amorphous phase behavior of a range of fullerene- and NFA-based systems has been determined and the metastable miscibility gap (the phase boundary) could be parameterized and thus quantified by the temperature-dependent Flory-Huggins interaction parameter χe(T). This has allowed to clearly show that there are two regimes: i) low χ systems with shallow quenches, where the best performance is achieved when the mixed domains reach the miscibility limit, and ii) high χ systems with deep quenches/low miscibility were the mixed domains have to be quenched into a composition that is close to the percolation threshold for electron transport. The deep quench systems are intrinsically unstable and further purification of the mixed domains often leads to burn-in degradation, although some highly vitrified systems have been discovered. We will review these developments and arguments and present some preliminary data that also hints that there might be a relationship between χ and the propensity for a system to vitrify and thus be stabilized.