They have some incentive to drive users to subscribe, but they have other forms of income, and I think if they ever implemented even a single feature of actual rendering that was closed source their community would riot and we'd get a community managed fork (probably by the guy who does the language server...).
The only way they can continue to gain traction is if they never ever in any way lock people to the web app. Documents must be portable, it's part of why someone would want typst anyways.
I do not see a future where this happens, and if it does it will be because the typst org has changed hands and is also no longer particularly relevant to the future of typst the language.
Is there really a community of volunteer contributors that could fork it if that happened? Typically with a corporate-backed project like this, the corporate development tends to crowd out the formation of a volunteer community of contributors that would be able to take over development.
All the typesetting extensions and such are a community effort. There are so many specific use cases that can only be/will be done by very specialized academics that a non-networked product would die on the vine.