I haven't read the Thompson but I have read the Althusser lately and understand it better. Logic vs. history is important, and it shows up most clearly in the moments when Marx says things like "so it is presupposed that somehow labor was separated from the means of production" but he refuses to speculate on how that happened. That is of course the lacuna for primitive accumulation and the historical backdrop to these economic models.
I think the more one reads historical work, the more one can pick fights with Marx's arguments about the spread of capital and the function of capital throughout the world. It almost seems like one could use two highlighters to break down the text: blue for capital's logic and green for capital's history. And you could contest the green sections while fully accepting the blue. Or use the green to contest the blue. Is this a dishonest mode of reading? At any rate, I agree that the division between history and logic is, unsurprisingly, well and alive.
I think that this distinction is perhaps what lies behind, for example, the recent call by Geoff Eley for looking closer at histories of capitalism. Namely, he argues that Marx and most Marxists presume capital's pure form to be free labor in urban settings. But if you look at the historical record of the last two centuries, those instances are a minority, at best (sorry for what I'm about to do):
Once we revise our understanding of the early histories of capital accumulation by acknowledging the generative contributions of slavery and servitude, in fact, we have already begun questioning the presumed centrality of waged work in manufacturing, extractive and other forms of modern industry for the overall narrative of the rise of capitalism. By shifting the perspective in that way, we effectively relativize wage labour’s place in the social histories of working-class formation and open our accounts of the latter to other regimes of labour. By that logic, the claim of waged work to analytical precedence in the developmental histories of capitalism no longer seems secure. As it happens, in fact, the de-skilling, de-unionizing, de-benefiting, and de-nationalizing of labour via the processes of metropolitan deindustrialization and transnationalized capitalist restructuring in our own time have also been undermining that claim from the opposite end of the chronology, namely from a vantage-point in the present. Today the social relations of work are being drastically transformed in the direction of the new low-wage, semi-legal, and deregulated labour markets of a mainly service-based economy increasingly organized in complex transnational ways. In light of that radical reproletarianizing of labour under today’s advanced capitalism, I want to argue, the preceding prevalence of socially valued forms of organized labour established after 1945, which postwar social democrats hoped so confidently could become normative, re-emerges as an extremely transitory phenomenon. The life of that recently defeated redistributive social- democratic vision of the humanizing of capitalism becomes revealed as an extremely finite and exceptional project, indeed as one that was mainly confined to the period between the postwar settlement after 1945 and its long and painful dismantling after the mid 1970s.
So what? One response to Eley has been: "Well if you had been an Asian/non-European historian for the last twenty years, you would've known this all along, genius."
But that is an unacceptable position too. Because the argument that Marx is too Eurocentric ignores the usefulness of Capitalfor understanding places that, although not a purely capitalist PRODUCERS, are nonetheless enmeshed in capital MARKETS (I made this distinction in my last post). And then combined and uneven development, etc. etc.