Despite the maturity of III-Nitride molecular beam epitaxy and the successful commercialization of III-Nitride products, many important details of the epitaxy of III-Nitrides remain unexplained. Effects of strain on growth kinetics, interface chemistry and electrostatics, polarization, and doping remain incompletely explained. This paper attempts to clarify some of the uncertain issues remaining in III-Nitride MBE while detailing new concepts such as polarization-engineered structures using polarization domains written into ferroelectric substrates. Topics to be addressed include the role of interface chemistry between oxide substrates (sapphire, zinc oxide, lithium gallate and lithium niobate) and III-Nitrides. The common role that oxygen plays in determining the structure of III-Nitrides near the interface will be examined as will the chemical dependence of and temperatures where oxygen is liberated from sapphire substrates. FET mobility can be varied from 46-1587cm@super 2@/V-sec in identical structures by varying the buffer layer nitridation temperature, buffer layer composition-either GaN or AlN at high or low temperatures, and buffer layer thickness. This variation is correlated to inversion domain, and to a lesser degree dislocation density as measured by electrostatic force microscopy. The use of near lattice-matched substrates supplies insight into the growth of GaN in the elastic strain regime. The surface reconstruction and surface smoothness in this regime differs from mismatched substrates and varies little with III/V ratio. Once the critical thickness is reached (~9 to 10 nm), the surface briefly roughens and further growth proceeds as with all other mismatched substrates. Finally, a new influence on film polarity is described, the control of polarity via electrostatic boundary conditions using ferroelectric substrates. Both potential applications and limitations of this approach for polarization-engineered structures is described.