I've been exposed to ODP or its precursor, Systemic Operational Design (SOD) mainly through reading and talking to SAMS grads who learned it in school. I think that it has great utility, especially for irregular warfare. One caution I would offer, however, is that even in its "basic form" it can appear overly complex and somewhat obtuse. That's not a reason to not pursue it, but I'm not sure we need ODP customized for various circumstances, complete with new definitions and concepts. ODP, by its general nature, should be adapted similar to the various military decisionmaking processes. As I read the current ODP, I think it can support SFA without alot of alteration.
What ODP highlights to me, and Rob captures it in his text, is that there isn't just "Blue" and "Red", but also "Green", the country/group we are trying to assist. ODP illuminates that we need to realize that our objectives are likely different than those we are working with and that our plan needs to take that into account. Understanding this up and down the chain, especially with advisors, would possibly relieve some of the frustration and bring better understanding of why the host nation security forces do what they do--their definition of victory is different than ours and thus their actions, all the way to the tactical level, will be shaped by their own objectives.
Additionally, ODP points out that we need as much intel collection directed at our "allies" as we do against the enemy. This places the advisor in an uncomfortable position. He cannot be seen as an intel agent by the very unit he is living with, but that information can be crucial to the intelligence community (and the commander) in developing the overall picture of the conflict. I have seen deliberate firewalls constructed between advisors and intel analysts charged with reviewing the host government.