Respectfully, I'm not sure what TreeVB means with respect to "completely trust CODIT". CODIT is a model for the concerted set of processes within the wood of living trees to resist the spread of the effects of injury and infection. Yep, some tree species invest a lot in compartmentalization, some not so much. And a starved individual tree may not perform as well as expected for the species in good health. Indeed, the CODIT walls buy time to restore lost sapwood function and the continuity of the vascular cambium. All trees die, all wood rots (in the greater conceptual fullness).
I do trust that CODIT provides a lot of explaining power. We don't know how all of the components work moment-to-moment, and there is species/environmental variation. But the relative strengths of W1-4 are pretty clear for most cases. Part of the challenge in teaching compartmentalization and CODIT is that Shigo used gross anatomical features for largely symplastic processes. If anyone is still reading, the sort-of-recent
article by Hugh Morris here is a pretty good treatment to which I provided a published commentary
here. Apologies, I've probably posted this before in TB, maybe even this thread!