This is a summary of the final chapter of the Essay, although the chapter is fairly short and hardly requires summarizing, so in addition to the summary, I think it would be fair to say a few words about the book as a whole.
The chapter discusses in a somewhat disjoined manner two topics: the question of the I and the issue of Maimon’s own position vis-a-vis materialism, idealism and dualism.
The section dedicated to “the I” presents a rather Kantian discussion of the basic difficulty of thinking the I – to put it bluntly (and simply): how can I think about the I if this very I is what does the thinking? Surely, I can think about thinking, but if the I is the ground of thinking (“a condition of all intuitions and concepts”), then my attempt to get at it is likely to fail since what I perceive/synthesize when I think about the I is already at work while I do what I do. To put it in Maimon’s terms, “as a result, it [the I, das Ich] can be thought as an object in general, but we do not have any cognition of it as a determined object (just because it is common to all objects).” [85/155-56] As Maimon illustrates this point, I can think of the I as a substance, but there is no way for me to have a cognition of the I as a substance because I have no intuition to subsume under it.
Let’s say I think that I am a cabbage laying in the field of cabbages that stretches as far as the eye can see. This scenario is both easy to imagine and at the same time impossible to think properly: it’s easy to imagine because even if I have never seen a field of cabbages, I’ve seen a field and a cabbage, so putting them together is easy; and yet I cannot be a cabbage, or rather, I cannot think that I am a cabbage, because cabbages do not think (as far as I can tell) and, most importantly, I don’t experience myself as a cabbage, because I have hands, feet and so on. Without going into much detail, we can say that either I am a cabbage that thinks it is a human being (in which case the question is how do I distinguish between the reality of being a vegetable and the illusion of thinking I am not one) or I am a human being that thinks it is a cabbage (in which case I have to explain how it is possible for me to think and be a cabbage at the same time). Before I lose you, let me reassure you that I am not insane, but I’m simply illustrating what Maimon will try to discuss in terms of materialism, idealism and dualism (and how it is connected to the I).
For example, the materialist (who gets a longer section than the other two), says Maimon, argues that, although there is seemingly a difference between inner perceptions (“inside”) and outer perceptions (“outside”), both are grounded in the same reality of matter. In this case, using my example, if I am indeed a cabbage in the field, I must explain why, even though everything is grounded in the same material reality, I appear to myself as a human being, that is, I have to account for the experience of myself as a non-cabbage, for the difference between what really (materially) is and what appears to be. One explanation is, of course, that I am insane, but then this explanation would require a point outside of myself from which I as a third person could evaluate my claim to be a cabbage. However, this third person position (I suppose one might read Maimon’s use of “transcendental object” as such) is impossible, so even if I am insane, I cannot ever know. Another approach can probe my statement “I am a cabbage” and ask me how I know that, or more precisely, how it is possible for my to claim that I am X to be true while I am so clearly and distinctly Y. I suppose one can imagine many other challenges to my odd assertion, but Maimon’s point is simple: “what the materialist understands by matter is the merely given, what exists in itself not by the operation of the power of representation. So the materialist claims that matter exists only in itself, not as a modification of the power of representation. For materialist, the I itself, or the power of representation, is a mere idea, and existence cannot be attributed to it.” [88/163] So if I represent myself to myself as a cabbage, my representation must be grounded in the outside reality of things and, unless I am able to explain how I can be mistaken in my self-representation, I am a cabbage. However, if I represent myself to myself as “the I” and also claim that I am a cabbage, I am either lying, because my self-representation is nothing but a synthesis of the manifold and is grounded in the same substance as the manifold (so really there is no “inside” or “outside” as heterogeneous spaces), or just extremely confused (or insane but in this case the only testimony I have is that of other human beings explaining to me that I am not a cabbage – empirical psychology).
The idealist and dualist accounts have their own issues which I am not interested in since they are similar to the ones posed to the materialist. Maimon’s own solution, presented later in the chapter, attempts to eliminate the need for “outside” (or “transcendental object”) altogether and I’m interested only in how it works, not in whether it’s true.
The example of cabbage is surely ludicrous and intended to illustrate a point – whether I believe myself to be a cabbage or “the I” (or a physical organism compiled of ever-so-small particles that no one can see but that are still posited as existing) the conclusion for Maimon is the same: “we can see from this that we cannot have a psychologia rationalis because we do not possess a concept of its domain that determines an object; but we can certainly have an empirical psychology.” [86/159] This empirical psychology can operate on the contingent level of my testimony (that I am a cabbage) or the testimony of others (that I am not), but it is not going to produce necessary and universal truths about “the I” (either mine or anyone else’s). Maimon’s own position is the following: if the object of thought is nothing but a manifold considered as a unity (and thinking is the act of synthesis), then taking distinguishing marks of an object together to form an intuition of that object (with help from concepts) does not require that any reference to anything outside of this intuition (“transcendental object”). The disagreement between materialist, idealist and dualist is the disagreement between the particular arrangement of two essential elements: the object and the representation of the object, the “outside” cause of intuition and the “inside” structuring and ordering that, putting various “distinguishing marks” together, gives us the object. I take Maimon’s proposal to be something along the lines of the following: we cannot with certainty (and therefore with necessity) establish the existence of the I that grounds our intuitions, either from a materialist or an idealist perspective (dualist perspective here is a kind of “having-it-both-ways” position that incorporates the problems of both), but we don’t need to, because we are falsely asked to choose between either complete certainty (necessity) or complete uncertainty (contingency) when in fact it is more than possible to imagine a kind of spectrum of reality where, by moving from ignorance to knowledge, we “attain ever greater reality.” [89/166]
Perverse Egalitarianism 
Mikhail: wow, those cabbages were really strange. But they did helpfully recast an old philosophical chestnut (the “brains in a vat” scenario).
Two questions on yr conclusion (&/or Maimon’s).
1– How is dualism (with its worst-of-both-worlds combo)really different from Maimon’s attempted escape from the “false dichotomy” of necessity and contingency?
2–The as it were asymptotic ascent to “ever greater certainty” seems to presuppose some vantage from beyond both the ascending line and the assumed goal. If I am getting ever closer to the goal without attaining it, in order to know this I need some way of gauging both my progress and the fact that my trajectory will never really get there. Isn’t this vantage the transcendental I again?
These aren’t objections necessarily, just clarifications.
Yes, cabbage in the field is brains in a vat etc etc, I just thought a new image might help liven up the experience. Now to see if I can address your questions:
1. Certainly, it seems that for Maimon dualism implies a kind of forever-split reality of this and that, whether it is body and mind or understanding and sensibility. I think, I might be wrong, that Maimon is moving in the Kantian direction of “humility of philosophy” in light of our limitations and suggests that Kant himself was not humble enough as he suggested we can both comfortably posit two realities and then also find a way to show how the gap (between intuitions and concepts) is breached. All we can say (this is more of how I see it, not being close to the text) is that the world as we see it (whether we are correct or mistaken) is the way we see it, i.e. some unified mass of different things that all fit together (identity and difference discussion in the Essay). Of course, this is just the starting point of our thinking (Hegel’s Sense-Perception chapter in PhG is one of the best expositions of this problem) and cannot serve as a foundation, so if we reflect on our immediate experience of the world, we propose certain models – and it’s not clear how dualism comes about in this scenario while the most immediate model would be that of monism (and then the next, a kind of infinitism as in “everything consists of an infinite amount of parts that themselves consist of part and yet it all somehow holds together”). To put it shortly, dualism is dogmatic because it is posited against the direct testimony of the senses and not enough evidence is provided in its support (argument goes, if we take Hume, from contingent experiences that forget that they are such and thus quietly become necessary/universal principles). In this sense, Maimon’s contingency and necessity are still on the same plane, still part of the same reality (unlike dualism where contingency comes from sense-experience and necessity from concepts).
2. In this sense, I would say that “certainty” has to be rethought in light of these new circumstances. If you are infinitely approaching the goal, you really don’t need to know how close you are to it, because you will never reach it and so your distance to it at the point X and as large (infinite) and at the point Y, even though you are closer to the goal. I’m pretty sure these ideas come from calculus for Maimon and as a non-mathematician I can’t say much, but only that there’s a goal against which you can gauge your success: infinite mind of God. How that works, I’m not sure as there isn’t much about it in the Essay, but I think Atlas has a good chapter on it, maybe I’ll read up later today.
hmmmm. If I follow you (or perhaps , if I run way ahead of you and off-track entirely), perhaps we can say that Maimon is groping towards a non-dogmatic equivalent of dualism. How very aufhebungen!