Suppose that we have $G(\overline{x},\overline{y})=-\frac{1}{4 \pi} \frac{1}{||\overline{x}-\overline{y}||}$ for $\overline{x}, \overline{y} \in \mathbb{R}^3$.
I want to calculate $\Delta{G(\overline{x}, \overline{y})}$.
I have found that $\frac{\partial^2}{\partial{x_1^2}} \left( -\frac{1}{4 \pi} \frac{1}{\sqrt{(x_1-y_1)^2+(x_2-y_2)^2+(x_3-y_3)^2}}\right)=\frac{1}{4 \pi} \frac{(x_2-y_2)^2+(x_3-y_3)^2-2(x_1-y_1)^2}{((x_1-y_1)^2+(x_2-y_2)^2+(x_3-y_3)^2)^{\frac{5}{2}}}$
Making the corresponding for $x_2$ and $x_3$ we get that $\Delta{G(\overline{x}, \overline{y})}=0$ for $\overline{x} \neq \overline{y}$.
Can we just say that $\Delta{G(\overline{x}, \overline{y})}=\delta(\overline{x}-\overline{y})$ if $\overline{x}=\overline{y}$ or do we have to show something in order to set that?
From what you have shown above, we know that $\Delta G$ is a distribution supported at the origin. To show that $\Delta G(x,y)=\delta(x-y)$, we can use its homogeneity of $-3$, $\frac1{\|x\|}$ has homogeneity $-1$ and $\Delta$ reduces homogeneity by $2$, and the Divergence Theorem: $$ \begin{align} \int_{B(0,r)}\Delta G(x,0)\,\mathrm{d}x &=\int_{\partial B(0,r)}n\cdot\nabla G(x,0)\,\mathrm{d}\sigma\\ &=\int_{\partial B(0,r)}\frac x{\|x\|}\cdot\frac x{4\pi\|x\|^3}\,\mathrm{d}\sigma\\ &=\int_{\partial B(0,r)}\frac1{4\pi\|x\|^2}\,\mathrm{d}\sigma\\[6pt] &=1\tag{1} \end{align} $$