There's an economic principal on how specialization can increase output for the same amount of work. Otherwise, clearly some failed startups can tell that they're doomed before they finally give up and this might give them something more valuable to do?
Will be interesting to see this strategy tested.