I did not dig deeper and I really don't think I need to. Average Prius owner (not including those fans in PC) will not agree that their city mileage is at lower 30's. When CR publish this, they should at least point out the difference between their result and most Prius owner's experience, and maybe further explain the reason. Note, the difference is at least 10mpg, not a small rounding error level of number. Maybe Bias is not the best word, but I have not find the better ones to describe this.
You should, to understand their test protocol. CR does not take surveys of what owners experience on any car in terms of fuel economy and it's not clear the results of that would be since there are SO MANY factors (e.g. trip length, speed, tire pressure, tires, habits, temps, bias from self-selecting audiences, how people determine their mileage [incorrectly, thru a trip computer [some of which are inaccurate], thru manual calculations, guesses,etc.]) and so on. FWIW, although I normally never visit Fuelly (due to the above), at http://www.fuelly.com/car/lexus/hs250h, I'm seeing 33.x mpg from a very tiny sample size, which isn't THAT far off from CR's 31 mpg overall on HS. http://www.fueleconomy.gov/feg/sbs.htm pegs '11 HS at 35 mpg combined. People can't even report their engine type correctly on Fuelly. Someone said their HS has an H4 engine. No, it doesn't have a flat 4. When looking at http://www.fuelly.com/car/chevrolet/cruze, I see some people reported their engines at Diesel H4 (again, they don't use flat 4s) or Diesel W8 (ha!).
You using the right gallon size? How long are you city drives in -4F weather? Care to indicate your worst mileage in that weather in your units so I can confirm?