I doubt that speculative realists have much to say about commodity fetishism at all, apart from the cited Harman essay, which has a defensive feel and a very restricted scope, for good reason.
I suppose there is nothing to keep OOP folks from leaving ontology at the office deciding to be Marxists, just as a post-Heideggerean like Sartre could adopt it in the political realm, though not out of any philosophical necessity. But from a Marxist standpoint I think Harman himself would come in for the sort of Marxist "scorn" he cites, as the kind of metaphysical wool-gatherer to which Marx said good riddance.
Unless one fences off commodity fetishism as a purely socio-economic concept, as Harman does, it seems to me that the two views are fundamentally incompatible. Or at least ships passing in the night and sinking out of view, beyond mutual rescue. Marx never, to my limited knowledge, spent much time on ontology and his "materialism" is a vexed and much-debated premise, in a way that wouldn't have much concerned Marx.
Yet despite his own denials or shrugs, I think Marx stands in the general penumbra of post-Kantian German Idealism, in which man is a self creating, unified subject-object, and in which knowledge itself is a compounding material creation. Just as we speak today of a "marketplace of ideas" one could expand the concept of "commodity fetishism" to the social production of ideas and philosophies.
In this sense, Newton's concept of gravity, for example, does not spring from the head of Newton but is the production of a given set of social relations. As is "object oriented philosophy," with the added epistemological absurdity that in this case minds produce "concepts" proclaiming their objective independence of all minds, a virgin nativity.
No matter what one makes of this, it is so far removed from the post-Kantian milieux Marx inhabited, even in his sweeping, unexamined "materialism," that it almost makes no difference. But I do think there is some justice to the charge of a species of "commodity fetishism" in OOP itself that Harmon tries to reject by narrowing the discussion.
I don't know a lot about speculative realism, and do find it somewhat interesting, but I think any claim to knowledge of unconditional noumena is patently incompatible with a view of knowledge itself as a social production, even a fetishised commodity of the linguistic exchange. So I think questioner is correct in suggesting this.