I think I have found an MJ/Bucklin multiwinner method that is more resistant to Hylland free riding than the usual Droop quota generalization of Bucklin is, and is relatively simple to explain from a constraint perspective. Would such a method be of interest?