Whether that strategy is a net win over a business cycle isn't known. Taleb's funds never published their full results.
I don't specialize in derivatives so I can't speak to how compelling his industry work is versus his writing. But my understanding is Taleb's strategies were explicitly designed to lose small amounts of money often and win huge amounts of money occasionally.
The idea is basically to go long vega and gamma waiting for an apparently rare event you believe will happen somewhat more frequently than expected. In the meantime you'll eat the theta and usually lose money, but ideally within certain risk parameters.
End result, even harder to win with OOTM options even if you are correct.
And yes it sometimes takes years to pay off, during which time the fund is paying management fees and options purchase prices. So it bleeds money over time, and then makes it back and more at random intervals.
That's by design. In fact it's a similar model to Venture Capital in that way.
It doesn’t matter if it bounces back later that year. Any puts near those levels would have cost peanuts and paid handsomely.