There's nothing indicating that the second or third opinion is correct.
Or will lead to a different outcome. The article says that in 21% of cases the diagnosis was completely changed, but not what the outcome of the treatment was. For a non-medical analogy, consider you have a rotting deck that needs fixing. You ask five builders in for a quote and get give different ways of addressing the problem. Most of them will probably end up fixing your deck, but they're all slightly different. Does this mean any of them are right or wrong?
An example from the medical field is blood pressure. Doctors are rewarded by their HMOs (see Goodhart's Law) for getting blood pressure within certain limits, so they aim for that even when it doesn't make sense - you get far, far more effect from lowering blood pressure at the extremes, even if you don't hit your target, compared to lowering it a few percent to hit your target.
So while the results are interesting, this needs more work to determine whether it's actually an issue or not.