(Abridged) The repeated discovery of supermassive black holes (SMBHs) at the centers of galactic bulges, and the discovery of relations between the SMBH mass (M) and the properties of these bulges, has been fundamental in directing our understanding of both galaxy and SMBH formation and evolution. However, there are still many questions surrounding the SMBH - galaxy relations. For example, are the scaling relations linear and constant throughout cosmic history, and do all SMBHs lie on the scaling relations? These questions can only be answered by further high quality direct M estimates from a wide range in redshift. In this paper we determine the observational requirements necessary to directly determine SMBH masses, across cosmological distances, using current M modeling techniques. We also discuss the SMBH detection abilities of future facilities. We find that if different M modeling techniques, using different spectral features, can be shown to be consistent, then both 30 m ground- and 16 m space-based telescopes will be able to sample M 1e9Msol across ~95% of cosmic history. However, we find that the abilities of ground-based telescopes critically depend on future advancements in adaptive optics systems; more limited AO systems will result in limited effective spatial resolutions, and forces observations towards the near-infrared where spectral features are weaker and more susceptible to sky features. Ground-based AO systems will always be constrained by relatively bright sky backgrounds and atmospheric transmission. The latter forces the use of multiple spectral features and dramatically impacts the SMBH detection efficiency. The most efficient way to advance our database of direct SMBH masses is therefore through the use of a large (16 m) space-based UVOIR telescope.