--and my understanding as to the distinctions and nomenclature is pretty much along the lines of that emedicine.com summary.

But--and this is an important caveat, gleaned not only from looking at a lot of the small-fiber skin biopsy material (especially the early work from the Johns Hopkins researchers) but from talking this over with Drs. Latov, Chin, et. al.--inflammation is likely a component of almost any of these immune-mediated neuropathies. Without getting into heavy biochemistry, the distinction is basically between those neuropathies that, like CIDP, have more classic "cellular" (macrophage/lymphocyte) infiltration, and those, such as the ones associated with M-proteins, in which the attack is protein/antibody mediated. The processes are biochemically dissimilar--you get "key in the lcok" receptor site cross-reactivity in the latter--but the result is still inflammatory, although it may be harder for measures such as C-reactive protein to pick up the latter. One of the most common findings on skin biopsy, for those who have small-fiber neuropathy, is swelling of the axons, and that's an inflammatory antecedent.

The example of small-fiber gluten neuropathy is instructive--in those with gangliosides along their C-fibers that are structurally similar to the gliadin molecule, antibodies formed to gliadin cross-react with those gangliosides, with the resultant small-fiber axonal dysfunction. There is generally a period of swelling and excessive branching before the fibers discorporate.