[I]n a dispute over the interpretation and application of the Vienna Convention, the United States cannot impose its own view on its treaty partners, or establish themeasure of its own treaty compliance.The first statement is true: the second is not. We are bound to establish a level of compliance that is different from the interpretation given by the ICJ if the Constitution requires it. That is to say, and as Damrosch concedes, if the ICJ ruling conflicts with a provision of the Bill of Rights, we must not give it effect (although she thought differently for the structural provisions of the Constitution it seemed). And if the ICJ were to rule on a case implicating national security interests of the United States, must we give it effect? I hope we don't agree to this when we enter treaties. The short answer is that we do, in fact, establish the level of our compliance from baseline domestic federal law and policy reasons.
So what should the level of compliance with the treaty be in this case? As a practical matter, review and reconsideration doesn't seem to ask for a lot (I don't know though how it would be any different from the review he has already received through his multiple appeals except to add, what if he had a Mexican consular there). But the real question, as others have pointed out more eloquently than I could, is why should the ICJ decision in Avana should be able to direct us that the treaty means this and we must create a remedy for it (not to mention, probably displace a clearly valid - even if expedited - judgment of the United States Supreme Court, which held in Breard that a state procedural forfeit rule was applicable even against Article 36 of the Vienna Convention).
Another quick note: she made a point that Scalia wrote in Olympic Airways - which I previously excerpted since I had a feeling she would try to get it in there - that international judgments have persuasive effect in the interpretation of a treaty. It's always nice to quote someone whom you know is on the other side, but not if the quote doesn't actually tell the entire story. In Olympic the Supreme Court was forced to interpret the word "accident" from the Warsaw Convention for the first time and decided it differently than two other international courts did. He found the majority's analysis no more convincing than these two international courts and would have adopted their analysis. Medellin presents the entirely different case of being directed by an International Court - not a sister signatory - that the treaty means X, you must accept that it means X, even though you have said it probably means Y, and you must enforce X in your federal courts and afford this individual a remedy in federal court, even though your laws might not provide for it. There is no evidence that Scalia would feel the need to give direct effect to an ICJ ruling. The present case is simply not the same as the Olympic case.