This was the approach Anki (which recently shut down) took for robotics. I actually had a similar idea, but whenever I went to the drawing board to make robot toys I could never make anything that was both fun and affordable - using videogames as a benchmark it basically always lost. I think Anki did as close to an admirable job as you could at trying to sell robotic toys while developing more utilitarian robots (they were working on a chore bot but it was woefully inadequate), but they still failed pretty hard. I was always surprised by their ability to continue to raise large sums of money (they raised about a quarter billion in total).

There is a similar approach being taken in the eVTOL space, with Kitty Hawk trying to develop their Heaviside as an obvious toy for the rich (single seater, short range, "quiet"). I think people will refuse to allow so much noise pollution in the skies, not to mention the inherent question of safety as all these designs make serious compromises to flight stability in order to make them so lightweight and take off vertically.

There is merit to this approach though - you could certainly argue hypercars are toys for the super rich and eventually that technology trickles down to cheaper vehicles. But the economics around racing cars are much more complicated, including making an entire sport (and an extremely popular one at that) to justify the development costs. You could also argue Tesla did this, starting with the Roadster, then the Model S, and now the Model 3. Tesla's financials have been a source of great debate for a decade now though.

Whatever the next big thing is, I bet the media hasn't caught onto it yet. I don't think it's VR, AR, self-driving cars, general AI, or anything else that has already been woefully overhyped. It'll be something nobody was looking for and the right person stumbles upon it at the right place and time.