TS Spec doesn't match implementation re method signature definitions

TS Spec § 3.9.5 says:

A method signature of the form

f < T1, T2, ... > ( p1, p2, ... ) : R

is equivalent to the property declaration

f : { < T1, T2, ... > ( p1, p2, ... ) : R }

But this is not the case with --strictFunctionTypes:

interface IFoo1 {
    bar(baz?: number): void
}
class Foo1 implements IFoo1 {
    bar: (baz: number) => void = (n) => void 0;
}

interface IFoo2 {
    bar: { (baz?: number): void }
}
class Foo2 implements IFoo2 {
    bar: (baz: number) => void = (n) => void 0; // err: Property 'bar' in type 'Foo2' is not assignable to the same property in base type 'IFoo2'.
}

I understand that the method bivarience in general is intentional as per the PR:

With this PR we introduce a --strictFunctionTypes mode in which function type parameter positions are checked contravariantly instead of bivariantly. The stricter checking applies to all function types, except those originating in method or construcor declarations

But it would be nice if the spec didn't state the syntaxes were equivalent when they aren't.

Open doc/spec.md at §3.9.5 and compare its equivalence claim with the strictFunctionTypes exception described in PR 18654. Update the specification so the distinction between method declarations and function-valued properties is accurately documented, then review the examples in the issue to confirm the wording matches their behavior.