Lambda conversions with unclear return type and ov

This question already has an answer here:

Peculiar overload resolution with while (true) 1 answer

If I have a lambda such as () => { throw new Exception(); }, it's unclear whether it has a return type or not. Because of this, it can be (implicitly) converted to both Action and Func<object> (or any other Func<T>). This is because, according to §6.5 Anonymous function conversions of the C# 4 spec:

[A] delegate type D is compatible with an anonymous function F provided:

…

If D has a void return type and the body of F is a statement block, when […] the body of F is a valid statement block in which no return statement specifies an expression.

If D has a non-void return type and the body of F is a statement block, when […] the body of F is a valid statement block with a non-reachable end point in which each return statement specifies an expression that is implicitly convertible to the return type of D.

But if I have two overloads of a method, where one has a parameter of type Action and the other Func<object>, and I pass it the lambda from above, the Func<object> overload is used. Why? Which part of specification says that Func<object> is better than Action in this case?

I have looked at §7.5.3.2 Better function member, but that does not explain it.

I should've looked one section lower: §7.5.3.3 Better conversion from expression explains that:

Given an implicit conversion C₁ that converts from an expression E to a type T₁, and an implicit conversion C₂ that converts from an expression E to a type T₂, C₁ is a better conversion than C₂ if at least one of the following holds:

…

E is an anonymous function, T₁ is either a delegate type D₁ or an expression tree type Expression<D₁>, T₂ is either a delegate type D₂ or an expression tree type Expression<D₂> and one of the following holds:

…

D₁ and D₂ have identical parameter lists, and one of the following holds:

…

D₁ has a return type Y, and D₂ is void returning