No, it's more complicated than that.
All voting systems have the potential for spoiler effects (in the broadest sense of the term). That's a core and long-proven theorem in social choice theory. What's more relevant is how those actually play out under the conditions in which they're used. And it turns out that, while pathological cases are still mathematically possible, in practice, under the conditions that typically apply to our elections, RCV is actually less likely to produce these effects than other systems.
The idea that approval voting, STAR voting, or Condorcet voting is superior to RCV for this reason is a misconception based on decades-old research that is no longer current.
(Also, the website linked above is not a correct demonstration of the effect you linked, although I can see how the confusion happened).