Sure, some people may spend
most of their time on Facebook, but it doesn't mean it is
all of their time.
The same people who spend most of their time on Facebook would probably be very annoyed if all of a sudden their online banking would need to be done through Facebook with a Facebook-approved bank, or if they couldn't book airplane tickets on a third-party site or check the weather and so on.
That's the problem with internet.org; it's not that people wouldn't spend most of their time on Facebook, but it's that it makes it harder for non-Facebook entities to get access to these users, essentially building a Facebook-branded moat.
For example, say that building an ISP to reach these users has a fixed cost of x. This means that if that ISP has n users, the costs of building that ISP are spread out amongst those n users (ie. x/n); however, with internet.org, this would lower the demand for this ISP (because n will be smaller, due to some people not interested in having anything else but Facebook) and because the price of getting actual internet will be higher.
It's definitely a good move for Facebook, but creates large negative externalities for other people. A better move overall (but less so for Facebook), would be to give free access to the actual internet but let them access Facebook on its own merits. If Facebook offers such value to these users (and it definitely does for a lot of people), then there is no need to wall these users off of the rest of the internet.
Edit: formatting.