That's good advice. Unless everyone agreed to run a prewritten module, expecting the players to do too specific things and go to too specific places is ill-advised.
Pre-planning too far, even with collaborative players, can fail to provide what they want. Only as the campaign progresses it will become clear what the players gravitate towards and what are their dispositions.
The GM can keep some general ideas for the future events and potential conclusions, but fleshing them out before they are imminent will only lead to wasted effort and disappointment. Being able to think on your feet is very important for a GM too.
Conversely, players should understand that GMs also have ideas of what they want to see, so they should at least try to pick up on some cues.
Everyone should remember that the core of TTRPGs is collaborative storytelling.
Trust issues is right. I wouldn't want to play with a DM who does that. This is not just fictional cruelty, to take a player out of the final battle just to rub in how evil and clever the villain is just feels unsportsmanlike to me.
But I see there is a whole genre of posts that's all about suggesting awful things for D&D games because it makes for spicy social media content. Same goes for, say, the False Hydra, a monster that does not fit with the games mechanics and imposes a metagame threat that bypasses the characters abilities and resistances entirely. I hope it's all just memes at the end of the day.