Greetings, Mohan,
I think the argument needs to be made more precise. For example, you ask us to suppose that art is a human practice that is an adaption. Before more can be said, we need to know what you take art to be picking out. Presumably, when you say that "suppose that art is one (human practice)" you DO NOT ask us to suppose that ARTWORKS are human practices but rather that ART PRACTICE(S) is a human practice(s) (that is an adaptation). As such, I take it that the argument looks something like the following (very rough) argument:

1) Practice A is an art practice only if i) A is some human practice P and ii) P is an adaption.
2) If P is an adaption, then we can derive a function-attribution of the form: F is a function of P.
3) For all P, if a P is an A, then if F is a function of P, then F is a function of A. 
4) F is a function of A.   
5) If F is a function of A, then for all w, if w is the product of A, then F is a function of w.
6) w is an artwork only if w is the product of some art practice A.
7) If w is an artwork, then F is a function of w.
8) If F is a function of w, then if w cannot perform F, then w is to that extant deficient.
9) Any w that is the product of P and cannot perform F is to that extant a deficient product of that P.
10) If P is an A, then if w cannot perform F, then w is to that extant a deficient product of that A.
11) Any w that cannot perform F is to that extent a bad artwork.

The obvious flaw as I see it is premise 5). For example, we could easily imagine a case of a P for which merely engaging in the practice brings it about that the function is performed (rather than engaging with the products of that practice). That is, it doesn't follow that if F is the function of some practice P then F is also the function of the products of P. Consider the following crude example: Suppose that the function-attribution derived from the human practice of sport is the acquisition and refinement of hunting/martial skills. It surely doesn't follow that any football, tennis, cricket match that does not perform that function is thereby to that extent deficient qua sport. Moreover, even were it true of the practice, I would still find it at least prima facie absurd to think that this entails that any (let alone all) evaluative/critical norms constitutive of sport must be explainable fully or even in part in terms of the acquisition and refinement of hunting skills (e.g., that soccer is better than water polo qua sport to the extent that the skills acquired and refined in soccer are more useful with respect to hunting that those acquired and refined in water polo).

Likewise, taking art practices (assuming we can coherently and productively carve them out) to be adaptions doesn't itself entail or suggest anything about the constitutive critical norms governing their products (i.e., artworks). What makes an art practice good or bad qua practice (qua adaption) needn't have anything at all to do with what makes an artwork good or bad qua artwork, and so any account assuming implicitly or explicitly there to be such a connection simply conflates evaluative/critical norms for the practice (artworld institutions&practices) with those for product (artworks). At the very least, such a connection must be argued for rather than assumed.

That said, my own view is that establishing such a connection likely requires either abandoning philosophy for sociology or committing several rather nasty acts of art theoretical, semantic, and metaphysical violence. No doubt there is a fascinating evolutionary story to be told; however, to take it as anything more doesn't signal a leap from the arm chair into the lab but instead a substitution of arm-chair brow furrowing with arm-chair yarn spinning.