The premise of a 10x developer is not that they work at 10x the velocity of a 1x engineer. If the average velocity per engineer on a team is 10 story points per sprint, a 10x developer is not going to complete 100 story points in the same sprint. That's impossible.
Instead, the 10x engineer will increase the overall velocity and code quality of an entire team, which could manifest as an additional 90 story points of velocity in a given sprint (if you can objectively compare story point quality across sprints).
If you pair a decent PM with a team of 8 junior-to-average engineers, the PM will have a roadmap that they can track towards with some level of granularity but will likely not be able to guarantee quality of the product and achieve reasonable timelines.
If you add a single 10x engineer to that team, the entire team will step up because 1) they can learn from someone and are motivated to do a good job, 2) they become more intellectually stimulated because they have someone they can bounce high-level ideas off of and actually get good answers, and 3) there's an implicit trust of having someone who is so much better than you telling you how things should be done and expecting you to step up and do it.
My best experience was working on a team as the sole PM with 15 engineers, 3 of whom were considered 10x engineers. As a result, we brought in engineers from other parts of the company who were considered average at best, and turned them into high-performing "5x" engineers. We did that about a dozen times over the course of 2 years.