Suppose $a_1...a_m$ is some composite natural number with at least two different prime factors written in decimal notation.
Is there an infinite number of composites with at least two different prime factors $a_1...a_m=\prod_{k=1}^{r}p_k^{b_k}$ such that $\text{rds}_{10}(a_1...a_m)=\text{rds}_{10}(p_1^{b_1}+...+p_r^{b_r})$
Here, $\text{rds}_{10}$ is the repeated sum of digits in base $10$.
For example $\text{rds}_{10}(123456789)=1+2+3+4+5+6+7+8+9=45=4+5=9$
Yes, there are an infinite number of such composites. One example of a set of such numbers is
$$(3^2)(2^{b_2})(5^{b_3}) \tag{1}\label{eq1A}$$
where $b_2 = 6n_2, \; n_2 \in \mathbb{N}$ and $b_3 = 3 + 6n_3, \; n_3 \in \mathbb{N}$. This is because the product is a multiple of $9$, so the repeated sum of digits will always end up at $9$. Also, since
$$2^{b_2} \equiv 2^{6n_2} \equiv (2^6)^{n_2} \equiv (64)^{n_2} \equiv (1)^{n_2} \equiv 1 \pmod{9} \tag{2}\label{eq2A}$$
and
$$5^{b_3} \equiv 5^{3 + 6n_3} \equiv (125)(5^6)^{n_3} \equiv (8)(1)^{n_3} \equiv 8 \pmod{9} \tag{3}\label{eq3A}$$
you then have
$$9 + 2^{b_2} + 5^{b_3} \equiv 0 + 1 + 8 \equiv 0 \pmod{9} \tag{4}\label{eq4A}$$
so it's repeated sum of digits will also end being $9$.