The simple fact of the matter is that nobody really knows why, and it's likely that nobody ever will. Certainly, there is plausible speculation around the matter, but it is only speculation. Unless and until we discover a series of (inexplicably prescient) sociolinguistic interviews from the period, the only evidence we will have is for that fact that it happened. Nor is the transition quite complete; there are still dialects where thou (or at least tha) is current, if hanging on by a thread.
To say that English was in a state of flux during the period when this occurred would be an understatement. It would likely be more correct to say that English, as a "real" language rather than merely a collection of nearly mutually comprehensible colloquial dialects, was coalescing out of the æther. Over the course of a mere two hundred years, we went from a situation where Londoners could not make themselves understood in Kent and translators just couldn't get a break (since so many of the words they used to translate works into "English" would be novel to some readers no matter which term they used) to a situation where a more-or-less standard English was indisputably a High Language outside of academia, you had taken over completely, and a King of England would use going to as a marker of the future rather than in the sense of "hieing itself thence" (which was its only meaning, as far as anyone can tell from the evidence, scant decades before). Near the end of that period, we have Shakespeare using thou and you in plays, both in the first person singular subject form and in the familiar, seemingly at random.
Yes, we have written and printed matter from the period, but that doesn't help a whole lot—at about the same time, the bifurcation of the language into distinct oral and written styles was also emerging, which considerably reduces the reliability of writing as an indicator of language change "on the ground". It can probably be fairly said that the emergence of a "legitimate" English literature had nearly as much of an effect on spoken language as changes in the spoken language had on the emerging English literature. And that bring us to the little matter of typesetting: th was frequently (though not consistently) typeset using a y as a substitute for þ, which means that any given occurrence of "you" in print may very well have been "thou" in the manuscript—and if the reader didn't know which was intended, and there was a sense that "you is the new thou" in the air anyway, that could only help to speed the transition.
My belief, and it is only a personal belief, is that the answer is (d) All of the above. We have the polite plural, a language that went from a disjoint collection of esoteric dialects to a consolidated "Real Language" learned exoterically by a large number of people over a (relatively) short period of time, accelerated by a new literature (and a spreading literacy occasioned by vernacular Bibles and the Protestant movement) printed by people who wanted to save paper and ink when they could, and whatever other potential reasons you can think of as well. Throw them all into a pot, simmer for thirty minutes over medium heat, and you get a definitive change in a pronoun, one of the most intimate parts of a language.
