That's a fair goal, and I believe the answer breaks down like this:
You can work for a big company that serves billions of people, claiming that you are making an impact on the world, while in reality doing nothing.
On the other hand you can start your own company, perhaps only directly impacting a few hundred people, and still accomplish the goal of maximizing your change on the world (as that was your ceiling).
Parker is arguing that in many cases, the engineer can actually have a bigger impact at the bigger company (e.g. because it leverages economies of scale and having lots of startups fragmentizes the talent pool).
It's a self-serving argument. A bigger overall impact is more likely when there are tons and tons of smaller companies out there trying ideas (i.e. the chance of one of them hitting the next big thing). If all these engineers just go work at Facebook instead, Parker will feel the biggest impact in his stock value. Individual engineers might do better at FB than they would have failing at their own company, but on the whole the economy lost whatever some of them would have made.
I'm not saying that I believe Parker's argument. I brought it up because the GP was considering the consequences of what would happen if Parker's argument were true. The GP, assuming Parker's argument were true, seemed to think that the conclusion was in the engineers' best interests.
What I'm saying is that, if you believe Parker's argument, it isn't necessarily against an engineer's best interests.
You can work for a big company that serves billions of people, claiming that you are making an impact on the world, while in reality doing nothing.
On the other hand you can start your own company, perhaps only directly impacting a few hundred people, and still accomplish the goal of maximizing your change on the world (as that was your ceiling).