In the unstable surface boundary layer, extremes in optical turbulence structure are encountered that seldom occur in astronomical applications. This is especially true when concern is with regions very near the earth's surface. Quantities defining optical turbulence spectrums, e.g. inner scale, outer scale and Cn2, exhibit distinct vertical structures that might reasonably be expected to influence path dependent turbulence phenomena. Propagation statistics derived using path-indexed turbulence spectrums requiring profiles for turbulent inner and outer scales in addition to Cn2 may differ from those predicted using the Kolmogorov spectrum, even when path dependence on Cn2 is included in both cases. To study this problem in an engineering context, the well-known modified von Karman spectrum is used to characterize optical turbulence structure. Because of it's relevance to a wide range of optical turbulence related phenomena, the mutual coherence function is selected as the propagation statistic of interest. To evaluate the mutual coherence function, a weak fluctuation gaussian beam wave model based on direct, high fidelity numerical evaluation of defining integrals is employed. Estimates of the wave structure function and related mutual coherence function are then derived for propagation paths through designated optical turbulence regimes. The vertical structure of optical turbulence is estimated by a model based on similarity theory and the universal equilibrium principle. Behavior of the mutual coherence function for plane and spherical wave propagation may then be examined for near-horizontal linear propagation paths through optical turbulence regimes representing the realistic extremes in structure characteristic of the unstable surface boundary. For purposes of comparison, parallel calculations of the mutual coherence function are performed by extrapolating from commonly employed engineering estimates based on the Kolmogorov spectrum. Differences in the mutual coherence functions derived using these two techniques reflect differences in spectral form, vertical structure in turbulence profiles, and approximations to integral forms typical of engineering applications. Results have significant bearing on the fidelity of standard engineering models employed in the assessment of optical turbulence effects on system design and performance.