Not the original commenter, but in my view as a crusty old software engineer, the reason to learn how transistors is because they're interesting and the key to how all computation is implemented. Once you learn how transistors work, you can learn how logic gates work. And then how CPUs and memory work. Bit by bit, computers stop being magic boxes and start being something built from technologies you understand.

Now, for most software careers (including my own), that knowledge isn't directly practical. BUT, the most important things for a successful, fulfilling career are curiosity and a willingness to dive into subjects _without_ necessarily knowing how (or even if) that knowledge will yield practical applications. That drive is what's going to lead to insights and breakthroughs over your career that wouldn't happen if you just let abstractions be abstractions.