I'd add a bit to that, at least in terms of how I use the terms. Autoimmune is, most correctly, just what Dr. Edwards states, although many others, including me, use it in a more sloppy manner to include things w/o antibodies such as reactive arthritis or ankylosing spondylitis, but these seem to be fundamentally different and would be treated differently.
B-cells make antibodies, so depletion of B-cells is logical generally when antibodies are involved. It's possible, however, that B-cell depletion might have other, more complex effects due to the interconnected network of cytokines that connect all immune cells. Still, B-cell depletion is most likely to make sense and to be effective in antibody mediated autoimmunity. Rituximab depletes B-cells (CD20 B-cells). It thus reduces antibody production. If you are making antibodies against your own tissue, getting rid of some of the B-cells that do that is to your benefit, obviously. My point is that while rituximab's effectiveness against ME (if shown in the phase III trial - which I think likely) is suggestive of an antibody mediated autoimmune process, it's not proof, and other mechanism are possible.
Lastly, I would include in "immune dysregulation" other immune disorders, innate or acquired, that lead to an abnormal immune function. For example, AIDS is a form of immune dysregulation, but it is not autoimmune or even autoinflammatory (like AS, ReA, etc.) An inborn genetic error could also be immune dysregulatory in nature - for example, SCID, or severe combined immune deficiency, is a genetic illness in which a person cannot produce antibodies. In addition to dysregulatory, which is a broad term, it's also immune deficiency (these people cannot make antibodies). One can argue that autoimmune must be a form of immune dysregulation, as proper regulation of the immune system would not allow autoimmunity.
There is some ambiguity in some of the terms as most people use them, but Dr. Edwards is generally very precise in referring to autoimmune disorders as being only those with a self antibody.