If something is just your personal, no-pressure, hobby project, maybe don't convince people to implement it in production?
People have to take responsibility for their own decisions about what to use in production.
personally I believe this is dependent on the marketing. elm market itself as production ready and reliable. that should mean that existing projects will not be scrapped by design.
I mean, this is even worse than python 2/3
I would totally be on elm side if the project came with a "not ready for long term investment or production" with it.
even in the real world this makes a difference. if I tell you that you can freely camp for the winter on an empty lot and invite some friends for free I cannot just evict you because I found someone who would pay. (in the real case likely you would sign a contract where you renounce this right).
similarly a project that invite users to use it loses some rights in term of calling others entitled when they raise some expectations
The reason why Elm is fun to work with is that it's Evan sharing his hobby project with you. And that's it. There is no support, no long-term maintenance, no implied warranty, nothing. Still, it works for me :) and it is Open Source in the way that I can fix bugs myself, if I have to.
One of the reasons why I chose to not do open source myself is because of people like the author here. You give someone free source code for a useful tool and they'll come back and ask for you to fix their issues.
For example, I once patched a Heroku buildpack because I needed it to work. Due to how GitHub works, my fork was public. A while later, Heroku linked to it. And then I started getting a lot of messages from whiny entitled pricks. Apparently, they thought that it would be my duty to provide support to them. For free, of course, because it's open source...
I'm pretty sure Evan had a similar experience and that's why Elm comes with no support, no warranty, and no way for you to influence it.