"What's the error on your estimate of how hard what he's done is?"
I don't like to think of hard in terms of the time required, but rather in terms of the skills needed and the things that need to happen to make the project a success. If you break down any project into a PERT chart, the best projects are the ones where each circle is capable of being realized in multiple ways. Whereas if you look at Sam's PERT chart, there are clearly a number of steps that are a one shot deal, where if they don't work the whole thing falls apart. For example if you can't partner with any carriers then the project is dead, unless you go around the carriers in which case your risk (and cost of capital) goes up so your upside goes down.
"What are the alternative ways of making $10m that you casually assume to exist?"
There are lots of ways. For example, doing LBOs on your own you can easily make 2-3 a year. The only problem is that you need the skills to know how to do this, and there are only a few hundred people at most who can serve as mentors to teach you at any given time. Running a fast food franchise is fairly predictable way to make 3-5m, so is starting a conference. If you start an ad network you can count on making 10+ if you know what you're doing.
There are tons of very predictable ways to make money on your own, the problem is they require skills that you can only acquire by finding a mentor who knows what they're doing. (And finding good mentors is a hard to acquire skill on its own.)