Deployment-Fidelity Gaps in Decomposed Algorithm Selection

The study identifies the deployment-fidelity gap between partition-level scores and actual end-to-end utility in algorithm selection. It proves that high partition-level scores can fail to certify the performance of a deployable system if the within-family selector is suboptimal.