I don't think anything here is in conflict with the statement that Fortran can be faster than C in some cases, but yes you're right that technically that should be qualified as naive Fortran can be faster than naive C