I have finally had time to read the full Boumediene opinions... and cross-reference various assertions in them against historical records. (One of the advantages of having a few thousand books in one's home is that one can grab them at 0300 when unable to sleep.) The most critical counterfactual is this one — and it's important, because Scalia's entire dissent is ultimately based upon assuming its soundness and validity:
The game of bait-and-switch that today’s opinion plays upon the Nation’s Commander in Chief will make the war harder on us. It will almost certainly cause more Americans to be killed. That consequence would be tolerable if necessary to preserve a time-honored legal principle vital to our constitutional Republic. But it is this Court’s blatant abandonment of such a principle that produces the decision today.
Boumediene, slip op. at 111 (logical) (Scalia, J., dissenting).
The counterfactual is simple: Put up or shut up. More explicitly, please show me a single instance since the rise of the nation-state (I'll even be generous and throw in the First Thirty Years' War) of asymmetric warfare in which the larger combatant's level of ultimate success was substantially undermined by its policy of following both its own domestic law and the then-customary laws of armed conflict. I've been unable to discover a single instance, but I don't have a (quite) exhaustive database; I haven't even uncovered a single instance in symmetric warfare during that period (and for this, I think I do have an exhaustive database). The real problem here is that Scalia has neglected the "more Americans to be killed" caused by the easier recruitment of terrorists resulting from GITMO's continued operation... although that question was, in fact, discussed in substantial detail in after-action reports concerning the Phillipines after the Spanish-American War, among other conflicts.
And that's before getting to the questionable reliance on so-called "tested precedent" as unquestionable — a variety of dead-hand control that would, I think, irredeemably and personally offend Justice Scalia if he'd apply it to the abortion-rights question. After all, Plessy, Lochner, Olmstead... and Roe v. Wade... were/are all "tested precedents." Johnson v. Eisentrager, which Scalia et al. would elevate to status similar to that of McCulloch (which, BTW, has had several parts later overruled by effect...), was wrong on the day it was decided; it was the result of hysteria on the one hand and poor lawyering on the other.
The whole point of the heirarchical system of courts that we have is that one court — the Supreme Court — is not bound by its own precedent, or that of any lower court. It is certainly guided by precedent, and needs to show appropriate deference to precedent to maintain the continuing function of the hierarchy. What Scalia et al. seem to be arguing for is the virtual codification of prior precedent, similar to that "provided" by Blackstone's commentaries on the law. (Of course, if Scalia et al. had been subject to some of the more-virulent and less-acknowledged bigotry also enshrined in those commentaries...)
In short, putting a thumb on the scales in favor of precedent may well be justified; pinning the balance beam in place, though, is not. The one constant in life is change. That is, indeed, why Marshall's oft-quoted, if seldom-contemplated, aphorism that "it is a constitution which we are expounding" is the real bedrock of a constitutional democracy or democratic republic: The implicit recognition that circumstances can change, requiring different means of expressing and implementing the underlying values in that constitution.
Later this week: three-fifths of all other speech.