The author concedes that smart contracts will likely find their niche one day.
...but I think that they have already found a good one: dramatically increasing the difficulty of collusion in any m-of-n scheme. Technologies which rely on m-of-n such as Shamir's Secret Sharing and Threshold Proxy Re-encryption (disclaimer: I work at NuCypher) have never found widespread use because the ease of collusion among unknown third parties is easy enough to warrant using known third parties (and once you do that, the use cases for such schemes tend to evaporate).
But look at the upcoming cohort of solid blockchain projects: many of them use smart contracts for collusion resistance.
...but I think that they have already found a good one: dramatically increasing the difficulty of collusion in any m-of-n scheme. Technologies which rely on m-of-n such as Shamir's Secret Sharing and Threshold Proxy Re-encryption (disclaimer: I work at NuCypher) have never found widespread use because the ease of collusion among unknown third parties is easy enough to warrant using known third parties (and once you do that, the use cases for such schemes tend to evaporate).
But look at the upcoming cohort of solid blockchain projects: many of them use smart contracts for collusion resistance.