From reading secondary literature on this last section, I learned that Marx wrote the following chapter on "former presentations of the subject" long before he wrote his specific criticisms of Smith and Ricardo in Part II. In fact, at this point in the text, one begins to receive the clear impression that this is an unfinished, unedited manuscript. This becomes more evident in the next two chapters. At any rate, Marx's writing here is clear and can be followed, but the specific importance of each line is uneven and sometimes barely there at all. At the risk of doing violence to the text, let me try to speculate on the actual theoretical significance at stake here.
He begins with Quesnay, who is chronologically situated before Smith but is really Marx's foil for Smith. Quesnay was wrong about agriculture as the source of value, but he was correct in his interpretation of production as the heart of the system and the way in which constant capital reappears in a renewed form in the commodity. Smith correctly expands value from agriculture to all realms of production, but he errs by not understanding the distinction between the pairs A) fixed and circulating capital versus B) constant and variable capital.
Rather than going into the details of the distinction (and I missed a large part of it since I did not read part II) let me give you a table from Harvey's Limits to Capital.
As far as I understand it, the misunderstanding is no more complex than: 1) the two binaries are overlapping and hence are confusing (perhaps due to the way Smith focuses upon the immediate physical properties of capitals in motion, unable to distinguish abstractions like constant and variable capital), and 2) by not seeing constant capital, Smith misses key analytical points.
Those key analytical points are: 1) by not seeing constant, Smith cannot see variable capital, by not seeing variable capital, Smith does not understand how exploitation occurs in the movement of:
M - C (CC + VC) - C' (CC + VC + SV) - M'
2) By not understanding constant capital, Smith cannot explain where the means of production come from, which means he does not understand the division of production into two departments (in the next chapter) and he does not understand that some departments of production will require more or less turnover time than others. In other words, Smith sees every individual capitalist as an isolated cycle of fixed and circulating capital generating revenue, which is then split three ways as wages, profits and rent (439).
However, Smith is aware of the failures of his equations; that is, he is aware that his equations cannot explain how capital (constant capital, investments) occurs. That is why "he has to smuggle in a fourth element by an indirect route, namely the element of capital." Marx then demonstrates that if Smith is going to insist that part of the "revenue" can be accounted for as capital, then Smith begs the question of where the value of that capital came from. For Marx, it is value that has been transferred from constant capital to the commodity.
On 444, Marx argues that Smith's fixed and circulating capital, described in terms of im/mobility, really just points to the two departments of production. Those that produce means of production (machines, gadgets, etc.) and those that produce commodities for immediate consumption (food, clothing). Because the latter circulate among consumers, Smith sees them as circulating capital. What he really should have been analyzing was the process in which value is transferred from department I to department II, that is, from Smith's fixed to Smith's circulating capital.
On 449, Marx points out a particularly vulgar part of Smith's calculus, when he argues that a commodity's original source of value is its division into revenues. I'm not sure exactly what Smith was arguing, but Marx's point is clear: Smith has inverted the historical process of production such that he first sees the finished product, the commodity, and retroactively constructs exchange-value from there. It is the ultimate economics of appearances.
On 451, Marx points out that even within Smith's own theories, the lacuna left by surplus-value emerging from variable capital is clearly detectable.
On 456, Marx shows another grave consequence of Smith's theory: by arguing that the workers' wages are paid out as revenue resulting from the sale of the commodity, he has collapsed the whole process of exploitation and rendered it invisible. Remember, the crux of Volume I is that a mystification occurs at the time of the sale of labour-power. That is, the commodity labour-power is sold at its exchange-value (the wage) in exchange for the use-value of labour-power (the ability to add surplus-value to the means of production). This site of mystification is where the worker gets paid and not at the time of sale of the commodity made by the worker. If Smith were correct, a jump in prices for a product would imply a jump in wages for the worker. But that is of course never the case.
Another conclusion Marx continually makes is that, according to Smith's formula, simple reproduction is the only possibility. If the output is equal to the inputs -- if revenue is only equal to the money paid by the capitalist -- then expanded reproduction is ruled out (457).
On 462, a crazy paragraph that argues that within a regime of value premised upon expended labour-power, a serf in feudal times produces value much in the same way as the wage worker. Wow, this is interesting. I'll have to think about this for my own research..."The substance of value is and remains nothing but expended labour-power — labour independent of the specific, useful character of this expenditure. A serf for instance expends his labour-power for six days, labours for six days, and the fact of this expenditure as such is not altered by the circumstance that he may be working three days for himself, on his own field, and three days for his lord, on the field of the latter."
"In this a commodity produced by a capitalist does not differ in any way from that produced by an independent labourer or by communities of working-people or by slaves."
Finally, on 464, Marx concludes that by not describing the commodity's components as different sections of capital, Smith has tried to create a transhistorical theory of value and production. In this way, Smith's serious error is that for him "the various factors of the labour process .... appear from the start in the character masks of the era of capitalist production." In other words, a bourgeoisie economist unable to think beyond the forms of appearance of capital.