I think that the differend between Kant and Hegel is still crucial, even from the point of view of speculative realism.Basically, Kant posits a limit beyond which thought cannot go — we cannot know things in themselves. Hegel’s critique of Kant is that, since thought is positing the limit, thought must always already be able to see beyond the iimit; for Hegel, the idea of inaccessible things-in-themselves is bogus, because it is our thought that has first posited them *as* inaccessible. This is what Hegel says, and it is repeated time and time again by later Hegelians, e.g. by Zizek.
Now, it has long seemed to me that any modern philosophy needs to begin with a counter-defense of Kant against Hegel. In the terms of SR, evidently, Kant is a weak correlationist while Hegel is a strong correlationist. Hegel’s argument against Kant is very much a form of correlationist argument — it is basically Stove’s Gem — to affirm something is to posit it and thence to know it. Hegel is simply saying that, by thinking that there are things outside of our thought, we are thereby bringing them into our thought. This is precisely the argument that SR most strongly rejects.
In this sense, even though Kant’s transcendental argument, his Copernican turn, is the locus classicus and foundational statement of correlationism, I think that returning to Kant can give us a way out of correlationism. This is partly what I was trying to do in my Whitehead book (even though I wrote most of that book before I became aware of speculative realism). Once we have made this return, several moves are possible. We may argue that the correlation, as Kant establishes it, is itself contingent rather than necessary (this is what Meillassoux does). Or, we may make a new version of the transcendental argument, asking not what our minds must be like in order for the world to appear the way it does, but what the world itself must be like in order for it to be able to appear to us the way it does (this is what Roy Bhaskar does). Or, we may extend the Kantian argument to all entities: it isn’t just human beings or rational beings that encounter appearances which are different from things in themselves, but every entity encounters all other entities phenomenally only, without being able to reach those entities in themselves (this is basically Harman’s argument that all objects withdraw, and that there is an unbridgeable distinction between sensual objects and real objects). Or, we may rework the transcendental argument as a principle of productivity rather than of essences (this, as far as I can understand it, is what Schelling does, at least in Iain Hamilton Grant’s reading of Schelling). Or else, as I prefer — following Whitehead as I understand him — we can invert the order of the Critiques so that the 3rd critique comes first — becoming, as Whitehead put it, a critique of feeling, which makes the other critiques unnecessary — that is to say, aesthetics precedes cognition — we affect and are affected by other things aesthetically before we cognize those other things, and even (or especially) when we cannot cognize them adequately. We cannot *know* things in themselves, or things apart from their correlation with us; but we can, as Harman rightly suggests, allude to them, i.e. refer to them metaphorically or indirectly. And we can, as well, be aesthetically *moved* by them — indeed, this is the primordial mode ofactual contact among entities (and in saying this, I am espousing a Whiteheadian version of SR which differs from Harman’s object-oriented ontology).