There is a proof online that shows that all metric spaces are normal. The proof is as follows
However, it has the additional baggage of needing to show that $d(x,A)$ is continuous and $U,V$ are open.
Can we do make it even simpler using the following:
Let $C, D$ be disjoint closed sets in $X$, $d$ be the metric that generates topology on $X$.
Define the set to set distance $d(C,D) = \inf\{d(c,d)| c \in C, d \in D\}$.
Then take $r = d(C,D). r > 0$ since $C,D$ are disjoint.
Let $U = \bigcup_{x \in C} B_{\frac{r}{2}}(x), V = \bigcup_{y \in D} B_{\frac{r}{2}}(y)$ are disjoint open sets containing $C, D$
respectively.
Is this good enough as an alternative proof?

It’s not necessarily true that $d(C,D)>0$ for disjoint closed sets $C$ and $D$. As a counterexample, take $C$ to be the $x$-axis in $\Bbb R^2$ and $D$ to be the graph of the hyperbola $xy=1$; then $C$ and $D$ are closed, $C\cap D=\varnothing$, and $d(C,D)=0$.