Investigate fixing vs not fixing scale and location distribution parameters

[noahprime]

In Stefan Lange's paper, there is no mention of fixing certain parameters when fitting distributions. However, this happens in the code for non-normal distributions. We should know why this happens, especially if we want to introduce additional distribution options in the future.