The problem with your theory, in my opinion, is that you are simply not accounting for cause and effect properly. Its like theorizing that people stutter because they get nervous. Well sure, nervousness increases sturtering. But nearly everyone gets nervous with public speaking, for example, but most dont stutter. Stutters are going to less spontaneous because stuttering can be a punishing experience and they begin to fear speaking situations. The same person, had he not been born with a stutter, could have ended up more "spontaneous". So first, there is a stutter. Second, if you have a stutter, factors are there that make it worse. As an analogy to what you propose, take the fact that most of us do not stutter when alone, but we dont need a study about the benefits of being alone.