That's a good point, but! I wonder if it couldn't be more efficient overall, if you look at it globally. Is an equivalent reverse osmosis installation cheaper or more expensive to produce, in terms of embodied energy that goes into manufacturing and maintaining all equipment, including control equipment?

Also, I wonder which system would be easier to slap together in a garage out of leftover junk. Not everyone can rely on access to commercial-grade, turn-key solutions to a problem. A design that could be reasonably DYIed could be better in certain contexts, even if less efficient.