Given the equation: $$ \frac{1}{0.005 + 0.0001x} \cdot \frac{1}{y} = z $$
Is it possible to figure out the $x$ and $y$ values for a given $z$, and if the point is not on the curve find the closest points on the curve. If I plot for some values of $x$ and $y$ I get a plot like the one shown below. So given the $z$ value of $192$ is it possible for me to determine that my $x$ and $y$ points are $2$ and $1$, respectively (actually at $x=2$ and $y=1$ the $z$ value is $192.3077$, but I would like to determine the closest point that is on the curve). Likewise for a $z$ value of $200$ I would get the same point. I would like to do this for any $z$ value.

If you solve the given equation $$\frac{1}{0.005 + 0.0001 x} \frac{1}{y} = z$$ for $x$, you get $$x = \frac{10000}{y z} - 50 \tag{X}\label{X}$$ and if you solve it for $y$, you get $$y = \frac{10000}{(x + 50) z} \tag{Y}\label{Y}$$ Substituting a constant $z$ into $\eqref{X}$ or $\eqref{Y}$ gives you the level curve or contour line where $z$ is constant.
If we rewrite the given equation as $$f(x, y) = z = \frac{10000}{(50 + x)y}$$ we can tell that at $x = -50$, $y = 0$, $z$ is undefined, and around that point, $z$ is antisymmetric with respect to $x$ and $y$. For this reason, if you are interested in integer $x$, $y$, $z$, it is useful to switch to new coordinate $$\chi = x + 50 \quad \iff \quad x = \chi - 50$$ so that $$z = g(\chi, y) = \frac{10000}{\chi y}$$ This is obviously antisymmetric in both $\chi$ and $y$, $f(\chi,y) = -f(-\chi,y) = -f(\chi,-y) = f(-\chi,-y)$. Because $\chi \in \mathbb{Z} \iff x \in \mathbb{Z}$, this translation does not affect 'integerness' or rounding.
To find all integer points $(\chi, y)$ that may be within one Manhattan distance unit (i.e., $x$ or $y$ off by $\pm 1$) of the contour line $g(\chi, y) = z$, calculate $C_1 = \frac{\sqrt{10000}}{\lvert z \rvert} = \frac{100}{\lvert z \rvert}$ and $C_2 = \frac{10000}{z}$, and do two double loops:
First outer loop is $\chi = 1 \dots \lceil C_1 \rceil + 1$, and the inner loop is $y = \left\lfloor \frac{C_2}{\chi} \right\rfloor - 1 \dots \left\lceil \frac{C_2}{\chi} \right\rceil + 1$.
The second outer loop is $y = 1 \dots \lceil C_1 \rceil + 1$, and the inner loop is $\chi = \left\lfloor \frac{C_2}{y} \right\rfloor -1 \dots \left\lceil \frac{C_2}{y} \right\rceil + 1$.
checking all points $\bigl(\chi, y, g(\chi, y)\bigr)$. Or, going back to the original coordinates, checking all points $\bigl(x, y, f(x, y)\bigr)$,
First outer loop is $x = -49 \dots \lceil C_1 \rceil - 49$, and the inner loop is $y = \left\lfloor \frac{C_2}{x + 50} \right\rfloor - 1 \dots \left\lceil \frac{C_2}{x + 50} \right\rceil + 1$.
The second outer loop is $y = 1 \dots \lceil C_1 \rceil + 1$, and the inner loop is $x = \left\lfloor \frac{C_2}{y} \right\rfloor -51 \dots \left\lceil \frac{C_2}{y} \right\rceil - 49$.
There are surprisingly few points $(\chi, y)$ or $(x, y)$ within those loops, so even a brute-force search to look for points that are sufficiently "close" to the surface is quite feasible.
The question of exact (minimum) Euclidean distance between point $(x, y, z)$ and surface $z = \frac{10000}{x y}$ is best asked in a separate question; but note that even then, because of the symmetries, you can safely limit to 3D octant $x \gt 0$, $y \gt 0$, $z \gt 0$.