I’ll point out that Servo was always a research project — any thought of it replacing Gecko was an incredibly long-term stretch goal. It was _a lot_ more interested in “can we write a high-performance browser engine in Rust” (there’s a reason why it significantly predates Rust 1.0 — they were very much developed together!) and “can we implement parallelism in places where browsers historically have not” (because you have more guarantees of safety).

Chromium _has_ been replacing certain components with ones in safe languages, and it seems like that will continue slowly (with integrating them into the existing codebase, with web compatibility risk of any changes in behaviour, with the risk of performance and power regressions).