You are contradicting yourself, or at least inverting the problem. Giving individuals control on how they can reuse their personal data maximises benefits out of the non-rivalrous property of personal data. No one suggested this would allow individuals to refuse inclusion of their personal data in national statistics, epidemiological research or road traffic optimisations. On the other hand, even vital medical research or smart city projects should still carry strong obligations towards the data subject. Sure, this might run against Open Data instincts, but it is only up to Open Data advocates to construct better anonymization systems. Some of those could, for instance, be built around the MyData notion, with differential privacy baked in.