I thought this would be good to discuss:
When Akira Toriyama was starting the Cell Saga of Dragon Ball Z, he never thought of Androids 17, 18 and 16 or even Cell himself until after Androids 20 and 19 appeared. His editors disliked 20 and 19 and asked him to replace them with "better" bad guys. As such, Toriyama had 19 killed quickly by Vegeta which led to 20 (Dr. Gero) returning to his lab to activate 17 and 18. The two cyborgs killed Gero and activated 16 before heading off to search and destroy Goku. Sadly, there's no pleasing some people and Toriyama's editors asked for 17 and 18 to be replaced. And so, Cell was created to absorb 17 and 18 and steal the show from them. Later on, Toriyama was asked to have Cell reach his Perfect Form (Kanzentai or Complete Form to be more accurate) faster since his editors thought his second form was weird-looking.
Where am I getting at with all this? Well, with all this knowledge in mind, I'm very impressed as to how Toriyama was able to write such a neat arc with so many changes to make along the way. Rather than canning an idea and instantly replacing it with another, he wrote the story to build up to the new idea and pass them off as plot twists which worked (most of the time). The new introductions don't seem at all asspull-ish and feel as though they were coming this whole time. Unlike a lot of cases of executive meddling, these changes manage to work.
But enough of Dragon Ball Z, let's apply this to writing in general. How would you change things when you're writing the story and you think that something different would be better than what you have now? Like a review gets you think or it dawns on you one day. Sure, with fan fiction, we can go back and correct stuff but what if you couldn't and had to change stuff up like with a published novel series (in the unlikely event, of course)? This is kinda the case with pantsers who don't exactly plan far ahead (you could call Toriyama one too).