The author did allege that the implementation took a long time _because_ the Go type system is "untraditional", which made the task at hand challenging. While

The author did allege that the implementation took a long time _because_ the Go type system is "untraditional", which made the task at hand challenging. While it's not clear what one might mean by a "traditional" type system, I'm claiming that the solutions to the problem of implementing generics in Go were well-known (not novel), and therefore the difficulty of coming up with them doesn't justify the long time it took to develop the feature.