There are plenty of editors which can be considered FOSS even by the pickiest standards and have Julia support, eg Emacs and Vim. They have a vibrant community of mostly individual developers, and based on their history, they are probably here to stay in some form in the long run.
Complaining that a mostly company-supported product doesn’t conform to the notions of FOSS that would be acceptable by RMS is a somewhat pointless exercise.
To be precise, the strategy stated by Juno developers is that they will work on the vscode extension - which as already noted, can be used with other editors, not only with VS Code.
That being said, it’s true that the developers of that extension seem to prioritise that the extension works correctly on VS Code, not that much on other tools. But I think that this is the right thing to do in the present. In this moment VS Code is by far the most popular code editor, according to the 2019 Stack Overflow survey, and this is a very good, pragmatic reason to choose it as the target platform for Juno, since that makes the life of the majority of users easier, and also attracts more people who can contribute to improve it.
Regarding the longer term, who knows which tool should be the chosen one? It may be VS Code as well for many years. If things start to go wrong with VS Code for whatever reason, it may be wise to choose a new editor to work with, as happened with Atom, but the most suitable replacement cannot be chosen with the information of today. And the amazingly quick move from Atom to VS Code proves that the strategy of basing the development of IDEs on independent building blocks is a good remedy for such unforeseeable problems, so there is no urgent need to anticipate them.