asks an excellent question: "What is it in particular about Gitmo that courts can't handle?"
I don't expect judges to be very good at knowing who is or is not a terrorist because I doubt the evidence the government has resembles the kind of evidence judges are used to seeing. Judges are used to seeing particular kinds of evidence in particular kinds of hearings. They generally hear live testimony from live witnesses, evidence in sworn affidavits, and the like, and they apply familiar legal standards to the facts to reach a decision. Of course they make errors in those proceedings, but at least the type of evidence and the context is part of the familiar day-to-day experience of litigators and judges.
My sense is that the detainees were mostly seized in battle or were handed over based on hearsay by allied forces. In that setting, hard and clear evidence of the sort that judges are used to working with is likely to be relatively light, regardless of whether an individual is a hardened al-Qaida terrorist or just some unfortunate fellow who was in the wrong place at the wrong time. Instead, the evidence is likely to be based on the impressions of agents about the reliability of third parties or fourth parties known years ago and not seen in ages, major parts of which will be classified. Judges are smart people, but I would think that this evidence is likely to seem pretty foreign and unfamiliar to judges used to run-of-the-mill competency hearings, probable cause hearings, sentencings, and the like.
Now, one response would be that if the government doesn't have the kind of strong evidence that the law usually requires to hold a person, then the detainees should be released. And if the judges end up releasing bad people who were held on very weak evidence, then that is the executive's fault rather than the court's . Maybe that's right. But if it's right, it's an argument about what the standard of judicial review should be, not whether judicial review is likely to lead to the release of bad people.
To be clear, I'm not saying that I think the institutional competence question should govern the issue. As I see it, whether the writ extends is distinct from whether judges are good at this stuff: Scalia's point about American lives being lost may or may not be right, but it shouldn't be relevant to the question of whether the habeas writ must extend to Guantanamo (although presumably it will be relevant to the subsequent determination of the proper standard of review down the road).