Hot take: There is way too little model reuse in computational social science. Lots of useful modeling frameworks but not enough collaborative model design, extension, and elaboration. Everything dies at version 0.1.0. Institutional structures compel shittier science.
I can. But, I think the mixture of roll-your-own and reuse favors the former way too much now. It be cool to see explandum-specific models that have the stature of OSS projects.