Well, the author seems to feel like he wanted to make a difference on the product side, and in that way the roads are not at all the same.

If your goal is to build software products, how do you make the biggest impact?

This feels like the question the author is answering with his observation that, as a programmer, you rarely have the clout to make the decisions necessary to drive the best product (or any at all) to market, outside of intentionally writing back code.