It hasn't been quite 4 years, but I have seen scant evidence these mythical, long-promised "significant further performance improvements every month for years to come" have materialized.

Then again, we're maybe only two years into the "4 to 9 years for optimizations" period.

On the gripping hand, MoarVM is now older than Parrot was when the Raku developers decided it was time to rewrite everything from scratch, so maybe it's time for them to start over again?

As I wrote years ago, I'm open to revising my opinion and estimation of Raku's current and potential future performance, given evidence, but I'm also going to hold previous predictions up to scrutiny.