Your link also discusses the later-no-harm criteria for Approval/Range leading to strategic voting, but all the examples I've seen for this seem to talk about giving a marginally preferred candidate an advantage in the absence of any truly disliked candidates. I think in practice, people will be far more interested in keeping disliked candidates out than furthering their marginal preferences.
Meanwhile, IRV suffers from Favourite Betrayal, which is a much more obvious incentive for strategic voting IMO. (Your link about NP-completeness of strategy only talks about STV.)