The analysis in this work essentially addresses the general question to what extent the temporal average of a particular quantity which is a highly nonlinear function of fluctuating quantities can be approximated by using the averages of the fluctuating quantities for its evaluation. The concrete case considered is the line emission intensity from sputtered impurities being a function of fluctuating electron density and temperature in a plasma beam of the PSI-2 device. A three-dimensional fluid model is employed to study the impact of plasma fluctuations on the distribution of particles and line emission in PSI-2 discharges and its interpretation in long-term measurements. In the model presented the solution of a vorticity equation to obtain a self-consistent electric field is avoided and a synthetic turbulent velocity field is included instead. This approach is based on a Langevin model including advection and allows numerically efficient parameter scans by controlling amplitude, correlation length and correlation time of plasma fluctuations known from extended 3D simulations and/or experiment. The synthetic turbulence model considered is an extension of established stochastic models used for studies of passive scalar advection and therefore, it is described in detail in a general framework. Numerical examples of PSI-2 applications show that a double log-normal probability density function for the electrons and impurity ions is likely to occur and that this supports the conclusion that very high levels of intermittency are required to find a significant impact on the experimental evaluation method which is based on temporal averages only. Consequently, for typical PSI-2 experiments the method of evaluation based on averaged plasma parameters is justified.