for a model to be called predictive, it has to be shown to be right once in awhile. Otherwise, it's actually just a hypothesis.
The scientific consensus against Einstein's theory of relativity thought they had predictive models, too. Read this and learn about the *actual* scientific method and what 'predictive' actually means in this context:
https://www.space.com/37018-solar-eclipse-proved-einstein-relativity-right.html