Another great post Joel. Couldn't agree more with your first paragraph when speaking about the "scale anticipation fallacy". Not only do people think they have to get everything being able to scale to millions, they also think in terms of "it only works if used by millions". Even worse if these two things go hand in hand. :S
I can also totally relate to the last part and have to say I am super proud to be as naive as I am, just like you wrote. Following my intuition as what works and what doesn't has paid of very well. Sometimes in the form of mistakes, sometimes as little successes. One thing for sure, they were and are always great points of learning.
Awesome post, let me Buffer this :).