My personal view, which I've yet to find time to write up as a a law review article, is that signing statements don't mean squat. Supporters of signing statements point to statements in two Supreme Court cases that I am inclined to regard as dicta (which is Latin for "stuff we lawyers ignore") in large part because the underlying theory of signing statements runs contrary to the generally accepted principle of delegation as set forth in Chevron, Mead and their more recent progeny. I am sympathetic to Scalia's argument that the Court has gone down a fairly inconsistent path on delegation, but he's been outvoted on this every time. So while I am inclined to think he has the better argument, it clearly is not the law.
Briefly, Chevron states holds that, when reviewing agency action, one looks to two factors. First, has Congress "spoken plainly" or is the statute "ambiguous" and therefore leaves room for agency interpretation. (This is the Chevron Stage I analysis.) If Congress has spoken plainly, agency action contrary to the delegating statute is unlawful. If a statute is ambiguous, however, a reviewing court must defer to an agency interpretation that is not "arbitrary and capricious."
Mead holds, as I read it, that a Congressional delegation is made to a specified agency based on that agency's specific expertise. In pertinent part, Mead states that when Congress does not specify an action, it is deliberately delegating authority to the expert agency because Congress intends that agency to act on the basis of its expertise. Accordingly, an interpretation by an authority not delegated explicit authority by Congress is merely "persuasive" and not binding.
Finally, even here, an agency is not free to disregard a statute because it considers it unconstitutional. Certainly an agency's opinion of the application of the statute in specific circumstances should avoid conflict with a constitutional provision. But where Congress has directly required an agency action, the agency is required to reconcile the directive of Congress with the Constitution rather than simply refuse to enforce a law. The sole exception is where relevant court precedent makes it plain that enforcement is always unconstitutional.
Under my reading of the relevant case law, the President is not an "expert" except in certain limited matters placed directly under his authority by the Constitution. This would include his role as Commander in Chief. His power to "recognize ambassadors" and negotiate treaties "with the advise and consent of the Senate" (which have been more broadly interpreted by the courts as the power to conduct foreign policy) and his power to appoint officers, with advice and consent of the Senate.
Bush and his supporters held that the theory of the "unitary executive" trumped the actual logic of delegation as set forth in Mead. They argued that because Article II of the Constitution makes all administrative agencies an extension of the President's general power under Article II to Execute the laws of the United States, it must follow that powers delegated to administrative agencies are delegated through the President and that the President therefore has the power to interpret a statute as if he were the expert agency to which Congress had delegated the authority in question. This form of the theory of the unitary executive is generally known as the "strong" unitary executive.
I argue this contradicts both the explicit logic of Mead and the language of the Constitution, which recognizes that the President appoints officers subject to certain constraints. e.g. The Senate must approve the appointment of "great" officers, may structure the mechanisms for the appointment of "lesser" officers, may set executive pay for these officers that the President may not increase or decrease, and may impeach these officers without regard to the will of the President. While the Separation of Powers doctrine likewise limits the power of Congress to restrain certain prerogatives of the President (such as the power to remove officers), it does not change the fact that the Constitution rec0gnizes the officers appointed by the President as separate from the President and having a separate legal existence and status that does not require the "unitary executive" theory.
Anyway, moving back to Obama and signing statements. According to the NYT piece, Obama has been issuing signing statements opining on the Constitutionality of various legislative provisions. In his case, to judge by the article, he has been sticking more closely to issues in the President's core constitutional powers (i.e., those powers explicitly assigned to the President, not generally provided under the "take care" clause) -- foreign policy and limitations on the power of Commander and Chief. Nevertheless, the process rankles members of Congress and others who thought that Bush's use of signing statements to interpret statutes or declare sections unconstitutional was a violation of the Constitutional limits on Presidential authority.
I'm inclined to agree that Obama ought to avoid signing statements that appear to interpret a law or find sections of it unconstitutional. Even if we accept that Mead is either inapplicable to powers explicitly assigned to the President or that in the operation of such powers the President is the "expert agency," a signing statement is an abstract and limited statement that gives every appearance of a line-item veto, a devices explicitly rejected by the Supreme Court as unconstitutional. Rather, IMO, the President should (a) limit himself to the process of issuing Executive Orders to other more usual means of addressing and clarifying certain provisions of law not delegated to a specific agency, or subject only to the President's general Executive authority, and (b) should, in all cases, seek to harmonize the statute with the Constitution. Only when a specific case arises that would clearly violate a constitutional provision should the President issue an Executive Order explaining his reasoning, in the manner of an agency issuing a decision, and subject to the same standards of review.
Obama is a constitutional law scholar. But he would do better to save his publishing his opinion on executive authority until after he is out of office, rather than in a signing statement.