This document does not intend to give an opinion on which fitting method is best, the idea of this document is to summarize the available ideas so the reader can have an informed opinion and is pointed to locations for further study. However there are a few general remarks we can make in summary.
There are a lot of different fitting options and none of them have proven to be superior in terms of performance (not even classical ‘all-electrode’ fitting). For setting T and C levels, live fitting methods seem to hold the biggest promise since they are quick and use a realistic signal for fitting. The combined data point towards the fact that for a group the exact setting of T-levels are not all that important and that individual electrode psychophysics testing is not per se better than any other method. Anecdotal evidence however, does indicate that for some individuals a precise setting of T and C levels can be important. There seems to be no obvious indicator at the moment to which recipients need manual fitting and which can be helped with fast fitting methods. This is an important research field for the future. There is no conclusive answer to the question whether or not using an NRT profile creates a better map, however, NRT certainly has its value in fitting since it can give guidance for initial map setting. Only one method has been tried so far (to my knowledge) that optimises parameters other than T and C levels. This genetic optimisation method may prove to be useful but we need more data to conclude this.
There seems to be some merit in self-fitting but it seems unlikely that we can ever create the perfect map from objective measures only; interaction with the recipients will remain needed.