Screened Coulomb potential in Debye-Huckel theory

I am rather confused about the meaning of the screened Coulomb potential in the Debye-Huckel theory. The potential of a unit charge inside an electrolyte takes the form
$$phi_{DH}(r) = frac{ e^{-kappa r}}{4pi epsilon r}$$
and satisfies the screened Poisson equation $(-nabla^2 + kappa^2)phi_{DH} = epsilon^{-1}delta(r)$. This also implies an exponentially screened charge-charge correlation function, i.e. $langlerho(r)rho(0)rangle sim e^{-kappa r}/r$.

However, I think $phi_{DH}$ does not represent a proper potential, since as‌ I explain in the following it cannot directly be used to compute the potential energy nor the forces inside the electrolyte. From this point of view, the screening effects would be contained in the correlation functions and the electrostatic interaction between charges remain unchanged.

Imagine we want to compute the electrostatic energy stored in the electrolyte. Using the Debye-Huckel potential, one gets
begin{equation}
U = frac{1}{2}int_{mathbf{r},mathbf{r}’} langlerho(mathbf{r})rho(mathbf{r}’)rangle , phi_{DH}(mathbf{r}-mathbf{r}’).
end{equation}
However, the screening effects of mobile charges in the electrolyte have already been taken into account in the charge correlation function; therefore, I think in calculating the energy we have to replace $phi_{DH}$ in the above expression with the unscreened Coulomb potential $1/(4piepsilon r)$.
The same question also applies to the computation of the body forces inside the electrolyte, that is I think the expression
begin{equation}
mathbf{f}(mathbf{r}) = int_{mathbf{r}’} langle rho(mathbf{r}) rho(mathbf{r}’) rangle left[-nablaphi_{DH}(mathbf{r}-mathbf{r}’)right]
end{equation}
is not correct and $phi_{DH}$ should be replaced with the unscreened potential.

Question is $phi_{DH}$ really just the charge-charge correlation function? can it be directly used to compute the energy or forces inside an electrolyte?