The same unchanged engine reads seven economies, and separates decision-grade readings from restatements of the input by testing every measure against a null.
Since September the engine also reads a published international index under the same discipline — proof the method is not tied to one subject.