P.S. A language that makes code reviews easier also works toward that goal of maintaining program correctness over time. The point is, there are many different

P.S. A language that makes code reviews easier also works toward that goal of maintaining program correctness over time. The point is, there are many different paths to correctness, all of them state-of-the-art yet are often in conflict with one another, and we don't have any mechanism other than empirical research to compare them. For example, is it beneficial to increase soundness at the expense of making code reviews harder? Not only do we not have an answer to that question, it is likely that there is no general answer (I say it's likely because whatever empirical research we do have shows messy results with large variance).