Clojurians Log v2

I do not intend to launch Electric v4 at all until our major performance problems are confirmed resolved. We do not feel it is important to match the state of the art frontend frameworks—Electric will always have more overhead due to needing to plan the network—but we would like to be comparable, so we are benchmarking early this time around.

Most of our historical perf problems are our own fault, it took us a while to understand the fine structure of what we were building and our implementation has contained shortcuts to get to working demos faster. Webassembly is not a magic performance boost because the hot spot becomes interop between the host language and the wasm kernel, but based on our early experiments it looks like it will be good enough. If it isn't then we will have to figure something else out, like use wasm only on the backend and have a js-native frontend for the dom rendering which is the heaviest part of our computation.