The $B_s to D_s^{(*)pm} K^mp$ decays allow a theoretically clean determination of $phi_s+gamma$, where $phi_s$ is the $B^0_s$-$bar B^0_s$ mixing phase and $gamma$ the usual angle of the unitarity triangle. A sizable $B_s$ decay width difference $DeltaGamma_s$ was recently established, which leads to subtleties in analyses of the $B_s to D_s^{(*)pm} K^mp$ branching ratios but also offers new untagged observables, which do not require a distinction between initially present $B^0_s$ or $bar B^0_s$ mesons. We clarify these effects and address recent measurements of the ratio of the $B_sto D_s^pm K^mp$, $B_sto D_s^pmpi^mp$ branching ratios. In anticipation of future LHCb analyses, we apply the SU(3) flavour symmetry of strong interactions to convert the $B$-factory data for $B_dto D^{(*)pm}pi^mp$, $B_dto D_s^{pm}pi^mp$ decays into predictions of the $B_s to D_s^{(*)pm} K^mp$ observables, and discuss strategies for the extraction of $phi_s+gamma$, with a special focus on untagged observables and the resolution of discrete ambiguities. Using our theoretical predictions as a guideline, we make simulations to estimate experimental sensitivities, and extrapolate to the end of the planned LHCb upgrade. We find that the interplay between the untagged observables, which are accessible thanks to the sizable $DeltaGamma_s$, and the mixing-induced CP asymmetries, which require tagging, will play the key role for the experimental determination of $phi_s+gamma$.