Re: [GHC] #14694: Incompleteness in the Coercible constraint solver (was: Can't coerce given assumptions)