We actually recognized that Skew had become a liability years ago, and Evan worked on a proof-of-concept to remove Skew around ~2020. But as described in the ar

We actually recognized that Skew had become a liability years ago, and Evan worked on a proof-of-concept to remove Skew around ~2020. But as described in the article, benchmarks showed that performance would have been severely impacted (mobile Safari was an especially big problem). The rewrite only became possible once mobile WASM was performant enough for us to move the hot parts of our code to it and we had enough engineering resources to do it safely.