Your comment poses a paradox. I strongly agree on some points and disagree just as strongly on others.
John Ehlers' very considerable reputation didn't come out of thin air. He has earned it by providing methodologies and software products that really work. His prominence in technical analysis is results-driven, by contrast with the marketing-driven profiles of many present day gurus.
On the other hand your view that one might make money by watching a market's supply-and-demand seems unpromising at best. Certainly we want to know "which way?" but what we really need to know is "when?" I'm curious about your "other data and instruments available to any trader with half a wit." If you refer to such traditional indicators as Relative Strength, Directional Movement, Williams %R and others that use arbitrary lookback periods I would advise looking elsewhere for something statistically valid that both avoids bad trades and indentifies promising ones. John Ehlers has developed and published several such statistics based tools that can be viewed on his website and in his articles, notably his application of the Hilbert Transform in identifying the current dominant cycle in any market and on any time frame. If you're unfamiliar with this aspect of his work one can understand your bafflement, but I'd urge you to explore his approach with a more open mind.
I think you err in saying, "...nothing as hard-wired as technical analysis will ever be able to automatically adapt, or therefore be useful, over any appreciable period of time." I completely agree with you if you refer to the traditional indicators that I mentioned earlier but I very much disagree with the notion that technical methods founded on the kind of applied science found in electronic signal analysis, for instance, have no value to the futures trader. First, technical analysis nowadays is anything but "hard wired" or inflexible. Second, few phenomena resemble market price movement more than signal behaviour.
Another ex-engineer who has created useful statistics based software is Cynthia Kase, author of "Trading With The Odds." The title speaks for itself. I believe it's long out of print but an ebook version exists and can be Googled. Her proprietary software suite, "StatWare," rewards attention also.
Again, I agree with you that no single technical indicator has all the answers. What is required is a realistic application of complementary tools driven by a top-down strategy that goes well beyond technical analysis. In my experience, for example, volatility is a core value while volume and open interest are all but useless.
With every good wish,
valleyvintner