1) I agree fully with Kevin. And let me know if I've misunderstood anyone here, but my understanding of the historic ICRC position is that once NIAC exists, it applies to all the territories of the parties to the conflict (whatever that may mean in the case of NSAGs). So when you refer to the ICRC's position re: "specific geographic locations" (rather than tracking the location of forces) it think that means all the territory of the party, not just the locale of hostilities. But perhaps more importantly, I think the ICRC would agree that it's statement was not meant to be all inclusive. For example, unlike Mary Ellen, and like me and (I think) Kevin, the ICRC would, I believe, agree that an Al Qaeda suicide bomber in a US shopping mall could be killed under IHL if all the underlying elements of humanity, distinction, proportionality, precautions, and military necessity are satisfied. The ICRC statement in your #5 does, however, raise questions about what the ICRC means to say on this issue. See my comment on #5, below.
2) It will be interesting to see how the U.S. treats targetability in light of Brennan's clear but incorrect statement that we could target all "members" of Al Qaida if we wanted to. His statement carries a patina of logic due to the popularity of applying IAC by analogy to NIAC. That notion is clearly wrong due to the difference between principle of distinction considerations applied to "armed forces" (whether in IAC or NIAC) and principle of distinction applied to NIAC against criminal enterprises like al Qaeda. The two categories of questions are 1) is it armed conflict? (apparently satisfied now in Syria) and 2) are the individuals targetable a) because they are members of armed forces of a party to the conflict (which I think is not greatly distinguishable from CCF), or b) because they are engaged in DPH at the moment of contact.
3) A nomenclature issue. You are using the term "internationalize" to refer to the nature and degree of interference from a State that changes a NIAC into IAC. The ICRC, contrarily, uses the term "internationalized armed conflict" to refer to a NIAC that remains NIAC, but that either a) physically crosses international boundaries or b) tht features a second State's support for the State party in an existing armed conflict between that State and a NSAG. Of course, if the second State "weighs in" on behalf of the rebels against State A, it's your use of the term "internationalize." To prevent this confusion, I think it better not to use the term "internationalize" when discussing the transmutation of NIAC into IAC, even though it's quite logical.
4) Fully agree that ICRC's views are not law. However, I would add that the ICRC's historic reticence to recognize and declare armed conflict, due to the decidedly un-humanitarian consequences of its application in contrast to peacetime law, means that by the time the ICRC gets around to saying "it's armed conflict," you can be pretty sure that it's armed conflict.
5) I think the only way to read this ICRC statement in a manner consistent with what I understand to be the ICRC's general position on the geographic scope of application of IHL in NIAC is this: "we think IHL applies to all the territories of parties to an armed conflict, regardless of the fact that hostilities may be limited to portions of their territory. But we know that this is not a universally shared view. Therefore, we remind you that even if you deny the application of IHL to some parts of the territory of parties to the conflict, you get the same results due to the continued application of IHRL."