Well, I got hold of a copy of German Science (in John Lyon's translation), so having re-read it, I can put a more precise characterization on my uneasiness at Ariew's characterization of the argument, and one that corrects a few details in my previous post. Essentially, Ariew identified the following characteristics of the argument of La science allemande:
1. The pursuit of truth requires both good sense, whereby we intuit principles, and logical reasoning, whereby we develop consequences.
2. Different minds are differently disposed to each; a mind more disposed to good sense (bon sens) is intuitive (un esprit de finesse), whereas a mind more disposed to logical reasoning is mathematical (un esprit de géométrie).
3. If good sense is allowed to dominate too much, we get the English mind, which is experimental to such an extent that logic suffers.
4. If logical reasoning is allowed to dominate too much, we get the German mind, which is logical to such an extent that good sense suffers.
5. The French mind is un esprit de géométrie corrected by good sense; it is therefore a balance between the two.
(1) and (2) are just Duhem's general Pascalian approach to science. Where I diverge from Ariew is in (3), (4), and (5). My reasons are the following:
(A) Duhem explicitly puts the French mind on the intuitive side. That is, the French mind is un esprit de finesse.
(B) The English mind is also on the intuitive side. It differs from the French mind in that the French mind shares with the German mind an interest in abstract logical structure. (This interest is expressed in different ways. The French, being intuitive, like to classify; the Germans, being geometrical, like to infer.) The English mind, on the other hand, is obsessed with models. This is because (as Ariew says) the English mind is far, far more imaginative than either the French mind or the German mind. This characteristic means that the English mind tends to be willing to accept reasoning that is in some sense broken and inconsistent, if it involves a vivid picture analogous to the thing being studied.
(C) Ariew is right that Duhem regards the English mind and the German mind as diametrical opposites. But his conclusion from this is not that they should both be eschewed, as one would expect if either were degenerate, but that they should get more exposure to each other in order to counter-balance each other. Maxwell discovers his equations by a series of inconsistent models and leaps of thought; Helmholtz starts working out a way to derive them rigorously. The French mind is in some sense between the two; but it is between the two in the sense that going too far in either direction is contrary to the French scientific genius.
(D) The French mind is not perfect scientific genius; even if it is in some sense between the English mind and the German mind, it is not a golden mean between them. Duhem is quite clear that the perfect scientific mind is the one that transcends national quirks, and exhibits the genius of all humanity. The minds of all the great scientists at their best approach this. The scientific mind only becomes national to the extent that scientific communities in those nations get off balance, deviating from the true. The English mind is too imagination-dependent and tolerant of contradictions; the German mind is too willing to sacrifice understanding to complicated reasoning; the French mind is too willing to flit from point to point and take things as obviously demonstrated that are merely plausible. On the other hand, the English mind is better than the others at taking everything into account; the German mind is better than the others at patiently following a line of reason through to the very end; and the French mind is better than the others at laying things out in such a way that the theory actually illuminates the phenomena and enlightens us as to their nature. They are each to be admired in their own ways, but they are each to be regarded with caution for reasons peculiar to them. Thus, all these different mentalities need to learn to live in harmony; the French mind is the indispensable complement of the German mind, and vice versa.
Duhem does hold that the intuitive mind is superior to the geometrical mind, and thus, to the extent that the German mind tends geometrical and the French mind intuitive, the French mind is superior to the German mind. But this is because science as conceived by Duhem must move from good sense to theory to good sense again, and the French mind is better equipped to determine what's good sense, just as the German mind is better able to work out the details of theories. The superiority is not one of the complete to the deficient, or the perfect to the degenerate, but of the end to the means. (How the English mind fits into this is never specified. But it, too, is intuitive. Its chief weakness is its dependence on models, which from the perspective of the French minds is an excessive use of a pedagogical crutch.)
How does Ariew get his interpretation? I suspect that Ariew is focusing chiefly on Lecture IV, where Duhem discusses all three minds, and treats the French mind as a sort of balance between the English and the German. But Lecture IV is chiefly concerned with giving young French scientists advice on how to work with their English and German counterparts. And the advice is roughly, that it's OK to be distinctively French in your manner of doing science (as long as you are striving for balance and recognize that the purest scientific work transcends nationality), that good English and German scientists are to be admired, but that this admiration should not lead French scientists to copy their vices. Thus by focusing too much on a few passages one could get the impression that Duhem thinks the French mind to be fully balanced, when he is actually advocating this as an ideal, not a characterization of the French mind. Likewise, Duhem is worried about the tendency of French scientists to get too mired in German obfuscation and thus leaving the virtue of clarity so assiduously cultivated by their predecessors; if you just focus on these passages the German mind doesn't come across too favorably. But there are other passages in which different purposes are pursued and you can begin to see more nuances to Duhem's position.