When performance matters, it's best to use the right tool.
If you require many joins and relationships across your schema, skip MongoDB despite your preference for a MEAN/MERN stack. This is what MongoDB says on their site:
NoSQL databases are designed to handle unstructured data (e.g., texts, social media posts, video, email) which makes up much of the data that exists today.
Multiple joins add up in MongoDB. You can read more about my opinion.