Very well put with respect to trolling. I, for one, appreciate the public dialogue. One nitpick: the law of large numbers doesn't mean what you intended here.
I hope the intention was clear: with 400 million users, it is a certainty that a feature will be incomprehensible, unusable, etc., to some non-zero number of users.
I'm now wondering what the rigorous way to describe the phenomenon you describe is. Something like, as the number of realizations of a random variable increases, the probability of seeing a realization below a given threshold approaches one. This sounds a lot like some of the theory related to hypothesis testing but it's been too long.
I hope the intention was clear: with 400 million users, it is a certainty that a feature will be incomprehensible, unusable, etc., to some non-zero number of users.