What would be more correct to write, i.e. if there is any difference, between:
$$ B=A(x,y,z)^{-1} \tag 1 $$
$$ B=A^{-1}(x,y,z) \tag 2 $$ Same question goes for matrix transpose, conjugate transpose, and other operations symbolized using exponents.
I think, though, that both may be confusing. Equation $(1)$ may be seen as the product of matrix $A$ with the inverse of a matrix formed by vectors $x, y, z$. Equation $(2)$ may be seen as the product of the inverse of matrix $A$ with a matrix formed by vectors $x, y, z$.
But supposedly we know that $A$ is a matrix function of variables $x,y,z$, then what would be the preferred notation?
As someone who is a matrix theorist, my preference would be for equation (1).
There are other options available. For instance, you could also write $$A_{xyz}^{-1}.$$
If the variables never change, you can make a convention that $A$ is a matrix-valued function depending on the variables $x$, $y$, and $z$ and then drop the variables à la $$ A = A(x,y,z)$$ and then simply refer to the matrix $A$.
Lastly, set $M= A(x,y,z)$. If this matrix is invertible, then you can simply refer to $M^{-1}$.
Although there are many-agreed upon notational conventions, often we're required to come up with our own on-the-fly. The key is deriving a notational system that minimizes ambiguity and maximizes clarity and neatness of exposition.