If the organisation you're trusting can't secure a single OS reliably, why on earth would you have confidence that it could vet every single app on its store and detect all possible abuses reliably? The latter is likely to be a much harder problem.