There's also a compounding factor. There's some relatively low-risk, high-reward career progression to be had by staying in something niche. I don't mean a JS packaging system, something more like A/B testing. People who stick it out advance, and niches are (by definition) small. So you're either bouncing around fields all the time (which can work) or the people you've worked with over the last 10 years are going to make up a surprising portion of the most influential people in your field.