I think this kind of catastrophic scenario can only happen if an app becomes completely discontinued, and no-one is able to upgrade it anymore. That happens, but currently the project is active and healthy and I plan to continue improving it for quite some time. I personally use it and there’s no alternative I like, so I’ve got a strong incentive to keep it working. There’s also more and more contributors familiar with the code base, which reduces the bus factor.
Also note that there’s the core functionalities of Joplin, like synchronisation, database, encryption, business logic, etc. are framework agnostics (the same code is shared by all the clients), so while creating a full fledge client would be hard, it’s not impossible as you wouldn’t have to re-implement the trickiest parts of the code base. I remember I was surprised how quickly I got a desktop version running, simply by re-using the core lib and mobile reducers.