use feather and pickle for faster loading
split x_ca processing into idca using chunking_size

and update comments
default behavior does not change, to enable use of feather
and the use of chunking, do this:

    # load from original
    model, data = component_model(modelname,
                                  return_data=True,
                                  alt_values_to_feather=True,
                                  chunking_size=chunking_size)

…into location_estimation_patch