Is there a "unique factorization theorem" for finite groups?

As CPM writes, its not true that $D_5$ is a direct product of $C_2$ and $C_5$, as I'm sure you're aware. Nonetheless, you may be interested in the directly indecomposable groups, which can be thought of as being "easy" groups. It is clear that for every finite group $G$, there exists a finite multiset of directly indecomposable groups whose product is $G$. There is also a uniqueness result, the Krull-Remak-Schmidt theorem.