Yes, absolutely. Gibbard's theorem is the starting point for this discussion. However, Gibbard's theorem doesn't say that voting decisions are always tactical, but rather that in any decision process (satisfying some very mild conditions), it is possible that voting decisions are tactical. It remains a legitimate goal to minimize the number of situations in which voters need to make tactical decisions to vote.

That's where this article comes in. With an ordinal voting system where voters rank the candidates, there is an obvious straightforward voting strategy where voters express their true preference order, and we can at least ask the question of whether that's the best way to achieve their goals, and if not, how exploitable the system is by tactical voters. Often with the best voting systems, it is optimal for voters to be honest; or even when it's not technically optimal, the depth of strategy needed to exploit honest voting is sophisticated enough that it's not feasible.

With approval or range voting, however, there's not a clear definition of what it means to cast a straightforward vote. You cannot even meaningfully ask the question of whether a voter should cast a straightforward ballot honestly expressing their preferences, because you're ultimately asking them a meaningless (or, to be more diplomatic, an underspecified) question, so the result will always depend on arbitrary choices they make about line-drawing or choice of units, as I explained here.