This is basically the inconclusive conclusion. Soy intake confers some benefit, but consumption needs to be regular and early in life. Just go to google scholar and put in soy, cancer. If you have institutional access you can read most of the articles.
The problem with all these types of questions is that the answers are essentially based on people’s recollection of what they ate months or years ago. How many servings of tofu did you have when you were 12? What’s a serving? Also, people don’t always know what they’re eating, foods interact with other foods in ways we don’t understand, etc. This is one of the main reasons that one week you read that pomegranate juice prevents cancer and the next week it causes diabetes. The smoking-lung cancer link was more complicated than most people think, but was much simpler than nutrition studies.