We have defined a methodology and format for capturing uncertainty in vulnerability curves, as described here in https://physrisk.readthedocs.io/en/latest/user_guide/vulnerability/vulnerability_uncertainty.html
This is also implemented in the class CDFBasedVulnerabilityFunction, which takes either a full cumulative density function (CDF) or standard deviations which can be used to create a CDF under different assumptions, and which can then create a conditional probability matrix via the prob_matrix_for_indicator_bins method. Module config_based_uncertainty_test.py contains tests of this class.
However, the class is not yet connected into the ConfigBasedVulnerabilityModelAcute class (only a stub exists). The work is to finish this plumbing and illustrate how that is used.