Correctly handle variances and correlations in powder workflow

[nvaytet]

Variances were dropped in the powder (POWGEN) workflow in #170 because broacast operations with variances are now forbidden in Scipp.
We need to implement an alternative way of keeping track of the variances (and correlations) in the workflow.
Should probably be based on an analytical computation of the correlations, as done in Heybrock et al. (2023) (Journal of Neutron Research).

[SimonHeybrock]

Should try the "upper bound" approach.