Your argument falls in the "nice gadget" category. Learning foreign languages ranks kind of high on the "privilege" ladder btw, it's very far from a "need", plus you don't need VR for that, we have video chats, youtube, instant translation tools, &c.
My argument is an example, from which one should be able to easily imagine other scenarios. A general class of "Remove artificial geographic constraints from knowledge work".
And our students do "need" foreign language training. This isn't something they're doing for funsies; their employers are paying for them to take the classes. These aren't executives looking to wine-and-dine business partners in China, these are diplomats, foreign aid workers, and intelligence analysts.
And these other methods aren't working to meet the demand for fluent speakers. You're not going to learn how to tell an old, former-Soviet base commander, "show me the nuclear storage facility access control logs now or expect some UN sanctions" from YouTube.
There is an international ranking schedule for foreign language acquisition, ILR Level, that roughly quantifies a person's fluency. Untrained, you start at 0. Most people get to level 1 by the time they finish college. Level 2 is basically competent enough to survive without a translator in-country. Level 3 is considered the minimum to be "competent enough to do work". There are a few higher levels, but everyone's goal is to get to 3.
With current training methods, almost anyone can be trained to level 2, 2.5, but there is still a big gap between what those people are capable of doing and where they need to be to do their jobs. We need more 3s and just doing "video chat" isn't working. Think, "the US military just sunk a cool million into your education to be able to translate intelligence on the fly, you better start showing results."
But there is some evidence that VR not only improves learning in general, but that it's particularly suited for language acquisition. UMD for example has a body of research they've been developing over the last 20+ years that shows there are real, psychological changes to people who perform language training in VR versus traditional and even flat screen simulations.
