It's definitely hard to change people, but I also think to some extent it's extremely easy if you know what you're doing and you're specific about the context. As the author notes: cults are really good at it. Another example is any entertainment performer, or even just a DJ: getting a crowd from one state ("cold") to another state ("hot") is well-understood game.
On the other hand scientists seem to be very bad at it (if the last few decades of quantitative social science research are any indicator). My pet hypothesis is that this is because an experimental environment is actually a very bad way to reason about human behavior. When you're dealing with basic material phenomena (particles, energy, chemical reactions, even biological reactions to a certain extent) the behavior is 'isolatable'. It will happen in a predictable way under controlled conditions, and removing outside influences makes that behavior easier to observe. Human behavior is the exact opposite of 'isolatable' though: so much of what makes us act the way we do is the influence of outside factors. In a general sense I don't even think it really makes sense to imagine a 'single person' or 'isolated behavior', our mind exists in mutual definition with the social world around us. Indeed the successful examples I listed above are successful precisely because they take control of your environment: a secluded meditation area or a sweaty club are critical tools for changing the way someone is.
As a result I think scientists often have a really skewed understanding of what motivates people, because anything they do in a controlled environment basically can never generalize by definition. Want to get kids to stop smoking? Don't treat the kids, treat the media environment they live in. But this isn't something that can be replicated in a generic/engineering/scientific manner; there's no magic recipe it's just common sense and coalition building (getting buy-in from the people who make media). But because these tools don't fit cleanly into an experimental framework they kind of get sidelined. Instead we get endless variations on nudge theory and little cognitive games to try and induce the behavior we want. Shockingly this fails time and time again. Preventing sexual harassment in the workplace simply cannot come from a PowerPoint deck or training, it will only follow from a massive overhaul in the cultural milleu. Look at any successful grassroots political movement: find people who are primed to receive your message; convince them to support you; use their support and evangelism to widen the circle to the next group of slightly less-primed people; repeat until massive; force structural/cultural change through a crush of bodies. Simple as.
edit: hm okay I clicked one link in the OP and it seems this point is not very original https://www.vitalcitynyc.org/articles/its-hard-to-change-peo...