While I do agree that producing thought-provoking technical demos has some (perhaps considerable) merit, the Media Lab still presents themselves as a research institute, not some creative agency making conceptual mockups/videos.
When something comes out of the Media Lab (or any other MIT-affiliated organization) the expectation will be that even if the ideas may not be workable immediately, the biggest technical/scientific hurdles have been overcome, i.e., they're not pure science fiction. MIT has built up reputation and goodwill so the public gives them benefit of the doubt.
Presenting half-baked, unworkable demos as research is to take advantage of (and exhaust) that goodwill, painstakingly cultivated over decades by the entire MIT community.