In the area of 1/4" playback one of the big issues has been issues associated with azimuth adjustment done automatically or perhaps better said to be optimized. Now - frankly, I do not know that everyone optimizes for each playback anyhow - but in a perfect world - if one had the time - one would.
One approach that has been taken in the Prestospace project is to develop a new playback head that is based on non-contact magneto optical reading - and there are actually many very tiny heads that are in the MO head. Because of the design the azimuth is essentially perfect - meaning that there is no azimuth error and the playback can be faster then real time.