I'm disappointed that this was posted with such a ridiculous assertion in the headline. Are you kidding me? Certainly the tests aren't entirely accurate, and I've complained about them, but saying that there's "No Valid Information At All" is bogus. Obviously you can go to the fueleconomy.gov site and see that there's a correlation between big, heavy, overpowered cars using lots of fuel and smaller, lighter, lower-powered cars that sip gas. The EPA has updated their tests a couple of times, most recently around 2007 following controversies that the Toyota Prius didn't achieve real-world fuel economy as good as what was on the window sticker. They also didn't try to factor in air conditioning or other features that are now common on cars.
The original 1970s-style tests produce numbers about 30% better than the end result today (an adjustment around 1985 reduced MPG numbers by about 15%, and the second one around 2007 brought it down by another 15%). Notably, government fuel economy tests in Europe and Japan still have ridiculously optimistic figures, so U.S. figures are much, much more accurate and reasonable compared to other places around the world.
Are EPA figures perfect? No. I personally think they went a bit too far in the most recent adjustment, since my (pre-dieselgate era) 2006 VW Jetta TDI gets MPG figures almost exactly matching what it originally had on the window sticker when I bought it.
And if this is all about people expecting to get super MPG when driving at 90 mph all the time, just stop complaining. That's not an appropriate expectation for what you should get out of these tests.