The proposal mentions that the compiler will prefer KeyPaths (or some subtype) over functions.
When inferring the type of a key path literal expression like \Root.value , the type checker will prefer KeyPath<Root, Value> or one of its subtypes, but will also allow (Root) -> Value . If it chooses (Root) -> Value , the compiler will generate a closure with semantics equivalent to capturing the key path and applying it to the Root argument
I think (I'm not an expert by any means) it would be difficult for the the compiler to know when to convert key paths into functions without any aide from us.
And as the very last sentence of SE-0249 says: " Accepting key paths directly would also be more limiting and prevent exploring the future directions of Callable or ExpressibleByKeyPathLiteral protocols". ("Callable" is what eventually became callAsFunction.) So the proposal is explicitly excluding this from working, because it would conflict with callAsFunction.