This is good.
I think you might want to go back to chapter 1, section of Steve's grammar, where he lays out what he means by "marked." He doesn't mean prominent. He's talking about marked for a feature (semantic, pragmatic, grammatical, syntactic, κτλ.). In the case of word order, he's talking about marked word order vs. a default work order. A marked word order is making something explicit that goes unmentioned in the default order (though perhaps still implied by the context). In the case of topical frames, there's nothing
inherently "prominent" about them, though that might be added by other factors. They simply involve the marking of the topic of the proposition which would otherwise merely be implied (or non-existent in certain clause types) and that the topic of this clause should be viewed as somehow distinct from the previous one.
That last bit is key, particularly here in John 10. The nature of information flow involves two factors.
First, its one-directional and generally linear.
Secondly, because of the limits of cognition, we're really only able to process a single proposition at a given time. Trying thinking two sentences at once. It doesn't really work.
The main exception to this involves forward pointing references (cataphoric pronouns) are the main one. But these involve distinct cues for the audience to recognize them. We'll come back to that...
Taken together, we have a situation where the processing of a given clause is first and foremost dependent upon what has come before. So an audience/reader encounters John 10:11a: Ἐγώ εἰμι ὁ ποιμὴν ὁ καλός. How do they process it? As it stand, they only have cognitive access to a few clauses back--you can only hold so much in your head as activated at one time. Every clause involves involves a proposition that consists of an assertion and (usually) an entity to which the assertion applies. The latter is the Topic of the clause. The former is the Focus of the clause. So under the model of information structure used by Steve (which was developed by Knud Lambrecht, btw), we have something like this for the information flow:
ὁ κλέπτης οὐκ ἔρχεται εἰ μὴ ἵνα κλέψῃ καὶ θύσῃ καὶ ἀπολέσῃ
In this first clause w have the topic ὁ κλέπτης and Jesus makes an assertion about what ὁ κλέπτης does (or does not do): οὐκ ἔρχεται εἰ μὴ ἵνα κλέψῃ καὶ θύσῃ καὶ ἀπολέσῃ
ἐγὼ ἦλθον ἵνα ζωὴν ἔχωσιν καὶ περισσὸν ἔχωσιν
The overt use of the nominative pronoun in the fronted position functions to mark the change of topic from ὁ κλέπτης to the speaker, who then make a parallel and contrasting assertion about himself.
Ἐγώ εἰμι ὁ ποιμὴν ὁ καλός
And here we are. We have an already activated referent (the speaker) being placed in the fronted position before the verb without any apparent change. It may very well be that this pronoun does actually, as you are suggesting, function in contrast to the forthcoming clause in verse 11. However, at this point in the flow of the text, the processing need of the audience is the evaluation of the relationship between this clause and the previous ones. The problem with suggesting that this is a forward pointing reference is that the audience needs explicit cues from the speaker to recognize such a reference. Demonstrative pronouns are the normal cue because demonstrative pronouns may involve ambiguities of reference. First person pronouns do not. They are, in fact, the least ambiguous pronoun a speaker could use (I'm me, after all. Who else could I be?)
But again, I'm not saying that there isn't a relationship between this clause and the later one. I'm saying that any such relationship that exists between them is not expressed until the later clause is spoke/written. The nature of the linear flow of speech requires that an audience must interpret the topic comment relationship of this clause first and foremost in relation to what has come before.
My own interpretation would be that Jesus is, as I said, making two parallel statements with ἐγὼ. The first functions as the primary contrast with ὁ κλέπτης. The second one allows for an additional assertion about who Jesus is with the same degree of significance placed upon it. Jesus could have said:
ἐγὼ ἦλθον ἵνα ζωὴν ἔχωσιν καὶ περισσὸν ἔχωσιν καὶ εἰμὶ ὁ ποιμὴν ὁ καλός
This, however, would have effectively subordinated the predication of Jesus as the good shepard to the predicate of Jesus coming. This makes little sense since the relationship between ἦλθον ἵνα ζωὴν ἔχωσιν and εἰμὶ ὁ ποιμὴν ὁ καλός is not a particularly natural one. And this fact point to the nature of the change marked by the fronted ἐγὼ at the beginning of verse 11. It is indeed highlighting a change, but the change that its highlighting is essentially a new paragraph. Jesus is essentially changing the subject. And a new subject needs a expression of the topic of predication, in this case: ἐγὼ. In this context, note especially what Steve says about asyndeton:
Runge, [i]Discourse Grammar[/i], 22-3 wrote:Asyndeton can be used at points of discontinuity, as at the beginning of a new thought or topic. Levinsohn summarizes the use of asyndeton in non-narrative by stating that since explicit connectives are used to indicate clause p 23 relationships such as strengthening, developmental, associative, or inferential, “the use of asyndeton tends to imply ‘not strengthening, not developmental, not associative, not inferential, etc.’ ” It is not only used in contexts where there is a change in topic (e.g., at the beginning of a new paragraph). Levinsohn notes that it may also be used in contexts of close connection, such as moving from generic to specific.
But it is also a necessary change. And I would say that you've intuitively caught on to something in your observation at this point: Ἐγώ εἰμι ὁ ποιμὴν ὁ καλός is a necessary clause for laying the foundation for the contrast that comes between ὁ ποιμὴν ὁ καλὸς and ὁ μισθωτὸς. However, the function of ἐγὼ is not in the contrast itself. Rather it provides the prequisite assertion that allows for the introduction of ὁ ποιμὴν ὁ καλὸς as a characterization of Jesus. Without this predication, the coming contrast doesn't make sense in the context.
ὁ ποιμὴν ὁ καλὸς τὴν ψυχὴν αὐτοῦ τίθησιν ὑπὲρ τῶν προβάτων
ὁ μισθωτὸς καὶ οὐκ ὢν ποιμήν οὗ οὐκ ἔστιν τὰ πρόβατα ἴδια θεωρεῖ τὸν λύκον ἐρχόμενον καὶ ἀφίησιν τὰ πρόβατα καὶ φεύγει
All of this is a fundamental result of the complex nature of information structure. Word order, conjunctions, deixis, and syntax all come into play and all contribute to the cognitive processing of the text on the part of the audience. It's incredibly complicated and requires the mastery of multiple fields of study that are only loosely related.