One limitation of the recently added --distance-matrix argument for the pathogen-cluster command is the lack of support for multiple input matrix files. If we wanted to support the same pattern of multiple alignment/distance matrix clustering for genetic distances that we support for embeddings, we would need to update the pathogen-cluster interface to allow nargs="+" and accordingly load and sum the distance matrices prior to clustering.
This is not currently a high priority but it came up during work on the corresponding manuscript.