While I think the concept of scalability in careers is very interesting, I'm not sure Taleb's "look at everything in the world in the context of my last book's title" framework provides the right perspective for the discussion. While the true home runs are indeed low probability events, risk/reward is quite asymmetrical in many scalable careers(i.e. there are no fat-tails to the downside). In the case of the author, the worst case is you spend a year writing a book, and receive zero revenue from the book. That is a pretty well defined worst-case, and is not particularly disastrous (assuming you can afford your expenses while writing).

Put differently, scalability does not imply high-risk. In the most literal sense though, non-scalability does imply limited potential reward.