It occurs to me there may be some misunderstadings going on between the con/pro camps.
If your goal as a designer is to trade max hp for mpg, a WAI is a dumb way to go. The better option is to stick a smaller engine in the vehicle. My calling WAI a "wacky" method didn't really express this properly.
Also, the argument that the sort of driving that hypermilers do may not be optimized for is fairly persuasive to me. One need not look further than the tire inflation specs to find evidence of where the sales priorities lie.
Assuming that a valve design that has bad flow resistance at low flow rates is the culprit (and that's not 100% certain), why would they not change the design? The aforementioned not testing that regime extensively is one, economic considerations of manufacture are another. There is another possibility. Now, this last one could get me in trouble with my fellow engineers, but don't assume the designers considered every possible implication of their designs. I went to school with some engineers who were definitely mediocre. There's a bias among many to credit professionals with a minimum level of proficiency that may not be warranted industry wide. The effect is very pronounced among doctors. I've known too many med students! In fact I recently discovered my freshman year roommate wound up in medicine. Yikes. The flaw may simply have slipped through, not being flushed out by a testing process aiming at different goals. Of course some cars may not have the flaw. That would make testing results variable.
None of this proves WAI definitively works of course.
__________________