React Native doesn't either. Only via the third-party React Native Web project[1]. Which I'm sure something similar for Flutter can also be achieved if desired.

[1] - https://github.com/necolas/react-native-web

While it does not right now and I doubt the team ever will do that, it is open source. I'm sure the team will accept meaningful contributions. A meaningful contribution here could be to make the flutter backend system pluggable enough that 3rd party backends can be used without modifying core. Then a web, desktop or any other backend can be implemented outside of it.

I know for a fast that the core team is open to the idea of creating desktop backends but does not have the resources to do it. They'll gladly accept any contributions towards that.

My point is the web frontend target is such an obvious one here. Specially given the dart to javascript compiler history.

I have very good experience writing UIs in HTML's canvas, look at my repos

https://github.com/netdur?tab=repositories

I say porting Flutter to web using canvas (both have similar APIs) is very bad idea.

this repo closes the gap in sane way https://github.com/yjbanov/butterfly

I don't understand what you mean. Please explain.

Adding web as a target on top of the enormous complexity of ios/android is a siren song. These cross-platform codebases are already held together with popsicle sticks and bubble gum with just the two targets.

Of note is that Appcelerator did target all three simultaneously (as well as blackberry and windows phone), but eventually dropped web support.

Good point. I develop all my apps using Cordova and it really is a big win that I can run these apps right in the browser with little to no effort. It's pretty trivial now to develop a Progress Web App and package it up with Cordova for some visiblity in the App/Play store.