Why does LINQ fail to scale when used as a universal data abstraction layer?

LINQ cannot uniformly abstract all data sources due to differing execution semantics and optimization models.

LINQ is designed as a unified query syntax, but underlying systems differ significantly: relational databases, in-memory collections, document stores, and remote APIs all have distinct execution semantics. This leads to mismatches in translation capability, performance characteristics, and supported operations. As systems scale, LINQ abstraction becomes a limiting factor because it hides these differences.