That question is up for debate, but I personally believe SoftBank was naive (and incompetent) but legit. The concept of blitzscaling is not entirely absurd. If you've identified a market with strong network effects, there will likely be a single winner. If you can pick one competitor early in the process, and give them all the money they need to crush the opposition in exchange for a hefty equity stake, and you don't, say, get confused and accidentally fund a bunch of their competitors, then you can probably ensure your chosen firm "wins", and the returns could be very positive. It's not clear that SoftBank managed to ever execute that strategy correctly, but it's not obviously crazy!