Feasibility assessment: wasm/WebGPU path for browser replay viewer?

Current engine/rendering code seems heavily tied to desktop OpenGL, while there are ongoing discussions about Vulkan/Metal/macOS renderer work (#2852, #2869, #2576).

I wonder whether a wasm build with WebGPU as a possible graphics backend could be worth considering as a long-term experimental target.

I do not mean full gameplay in the browser, or full LuaUI/widget compatibility, as an initial goal. I also understand that Recoil replays are not simple video recordings; accurate replay viewing may still require substantial engine simulation and game content support.

A narrower starting point could be browser-side replay tooling: metadata parsing, timeline/statistics extraction, or eventually a constrained replay viewer with built-in UI only. A first proof of concept could intentionally avoid rendering and focus on replay parsing or headless replay-derived data.

Would maintainers consider a small proof of concept in this direction useful, or is wasm/WebGPU too far outside current priorities?

Start by mapping the current engine/rendering code and replay-parsing path described in the issue, then assess whether a browser-side, non-rendering proof of concept is feasible. Done would be a maintainer-backed decision on a small wasm replay-tooling experiment, rather than full gameplay, LuaUI compatibility, or a complete viewer.

Written by the indexing model from the issue text.