Hi @CookieSakje,
It’s very difficult to tell what’s going on in these screenshots because they are too blurry to read the text, so I can only give you my best assumption. Based on what you describe, it sounds like you have some parameters that are set to a different value than the Megascans material, since it looks like yes, you have matched all of the nodes. When using the buffer visualization, what specific channels look incorrect?
If all of your values match, confirm that the material functions you are using are exactly the same.
Here’s a question though- why not just use the Megascans material? Or duplicate that material so you have your own version? This will probably be a lot cleaner and a lot less hassle.