Not yet reviewed by Pith; the record is open.

We consider the asymmetric random average process (ARAP) with continuous mass variables and parallel discrete time dynamics studied recently by Krug/Garcia and Rajesh/Majumdar [both Jrl. Stat. Phys. 99 (2000)]. The model is defined by an arbitrary state-independent fraction density function $\phi(r)$ with support on the unit interval. We examine the exactness of mean-field steady-state mass distributions in dependence of $\phi$ and identify as a conjecture based on high order calculations the class $\mathcal{M}$ of density functions yielding product measure solutions. Additionally the exact form of the associated mass distributions P(m) is derived. Using these results we show examplary the exactness of the mean-field ansatz for monomial fraction densities $\phi(r)=(n-1) r^{n-2}$ with $n \geq 2$. For verification we calculate the mass distributions P(m) explicitly and prove directly that product measure holds. Furthermore we show that even in cases where the steady state is not given by a product measure very accurate approximants can be found in the class $\mathcal{M}$.