We present theoretical continuum emission spectra (SED's), isophotal maps and line profiles for several models of photoevaporating disks at different orientations with respect to the observer. The hydrodynamic evolution of these models has been the topic of the two previous papers of this series. We discuss in detail the numerical scheme used for these diagnostic radiation transfer calculations. Our results are qualitatively compared to observed ultracompact Hii-regions (UCHii's). Our conclusion is that the high fraction of ``unresolved'' UCHii's from the catalogues of Wood & Churchwell (1989) and Kurtz et al. (1994) cannot be explained by disks around massive stars. In particular, the observed infrared spectra of these objects indicate dust temperatures which are about one order of magnitude lower than expected. We suggest that disks around close companions to OB stars may be necessary to resolve this inconsistency. Alternatively, strong stellar winds and radiative acceleration could remove disk material from the immediate vicinity of luminous O stars, whereas for the lower luminosity sources considered here this will not occur. We also find that line profiles tracing the evaporated material originating from the disk are not influenced significantly by the existence of stellar winds over a wide range of wind velocities (400 - 1000 km/s). We compare our results to the bright IRAS source MWC349A. Many of its properties, especially its spatial appearance in high-resolution radio maps, can be well explained by a disk surrounding a UV luminous star with a high velocity stellar wind.