Check out the book "How Markets Fail" for a discussion on how the efficient market hypothesis gained popularity (part 1) and a discussion of the many ways that real markets do not maintain the preconditions for markets to be efficient.

In particular, the precondition that Adam Smith stated was that increasing production required linearly increased costs. However, this is frequently not the case (utilities, railroads, software, social media networks, etc.). Also, markets require a certain minimum number of producers (and consumers) in order to remain efficient. Consolidation can reduce the number of producers below the level that meets the precondition of efficiency, and then barriers to entry to prevent newcomers from joining the market.

As far as efficient market theory applied to the stock market, I've always thought that efficient market theory was trivially disprovable: did all companies in the US lose 20% of their value all at the same time this past Nov - Dec, in the same season that they are selling all their inventory during the holiday season? And then regain 10% of it all at the same time the next month? Did Apple lose 50% of its value from between July and December, despite the fact that its revenues, costs, dividends, assets, people-capital, know-how, brand, consumer sentiment, etc. remained roughly the same? Compare to, say, PG&E during the same period.