Its not A or B; its A, in part because of failure to correctly anticipate B. Predicting market success is, inherently, a prediction of market participants successful execution.
I said this in another thread. We know that self-driving cars are gong to be a thing in the future. That doesn't necessarily mean that the effort a company like Google makes today is going to be that valuable. It might be that when full autonomy is ready for mass implementation, we already had cars running around with (much cheaper) sensors for years. Since things like highway autonomy and active safety systems are lower hanging fruit. On the other hand if they gather enough patents and full autonomy explodes in 10 years, they have a good position. There's all this complexity yet most of HN is like "Google is the self-driving company".
Which begs the question "were predictions too early, or were early executions insufficient". There are a million examples of first movers who didn't execute. Off the top of my head Yahoo, Blackberry, and IBM OS/2 were first to market for generational techs and then lost.