In this study experimental single-pulse, time-resolved laser-induced incandescence (TIRE-LII) signal intensity profiles acquired during transient Diesel combustion events at high pressure were processed. Experiments were performed between 0.6 and 7 MPa using a high-temperature high-pressure constant volume cell and a heavy-duty Diesel engine, respectively. Three currently available LII sub-model functions were investigated in their performance for extracting ensemble mean soot particle diameters using a least-squares fitting routine, and a "quick-fit" interpolation approach, respectively. In the calculations a particle size distribution as well as the temporal and spatial intensity profile of the heating laser was taken into account. For the poorly characterized sample environments of this work, some deficiencies in these state-of-the-art data evaluation procedures were revealed. Depending on the implemented model function, significant differences in the extracted particle size parameters are apparent. We also observe that the obtained "best-fit" size parameters in the fitting procedure are biased by the choice of their respective "first-guess" initial values. This behavior may be caused by the smooth temporal profile of the LII cooling curve, giving rise to shallow local minima on the multi-parameter least squares residuals, surface sampled during the regression analysis procedure. Knowledge of the gas phase temperature of the probed medium is considered important for obtaining unbiased size parameter information from TIRE-LII measurements.