if you are interested in building height, just use the standard PS approach, it’s the best tool you can find around for that, and it will give you all what you need.
Tomosar by itself would mean finding multiple scattering centers in the same resolution cell. the amount of additional infos you can find with respect to the standard approach is pretty limited, it is mostly a theoretical exercise. In Sarproz if you want you can do it by selecting e.g. 2 scatterers. However, after doing so, you have to filter the infos to find what you need. We never really use it.
Alternatively, you can implement in Matlab your own approach and use the data preprocessed by Sarproz.
Consider anyway that Sentinel is not a good sensor for height estimation because the normal baseline diversity is very limited