Well, I recently purchased a book called "Evidence Based Technical Analysis" (Thank you CV for recommending it) and it has changed my idea of price action,indicators,patterns etc completely. To what extent I am educated enough to write a review is questionable but anyway, here it goes:
Unlike the books you have mentioned above, it does not try to explain a trading system or try to detail if the market is "bullish" or "bearish" nor does it give any specific trading idea. The following areas are covered by the book:
---> Questioning the authority of subjective technical analysis (which includes EWs,Patterns and anything which can't be backtested, for want of evidence and serverly lacking statistical literacy)
---> An insight into theories of price motion (which I felt was the best part of the book)
---> An explanation of data mining bias (and not to mention the other biases explained eloberately)
The book concludes by testing 6400 strategies in an attempt to prove the predictive ability of technical analysis and disprove the random walk of EMH by earning a risk adjusted rate of return greater than the benchmark.
An excellent read and truly a book worth admiration. The bibliography alone went for 50pages!