Have you ever thought about how is the cross country speed measured? If not, think now! It's absurd to mix data from different sources as your web (cited by you) does. Or do you think that the test conditions where everywehere the same? The same moisture, the same vegetation, the same percentage of uneven terrain, perhaps even the very same terrain?
And what about factors like driver (the guy driving)? If you don't know it, it matters. Do you think the driver was the same all the time? These all are reasons, why is the crosscountry speed rarely mentioned and taken seriously. This is why the numbers like kW(hp)/ton or kg/cm2 (pressure on the ground), and so on, are used to give you a picture of crosscountry performance, rather than a single speed number got under unknown and unreplicable test(s). And because this is over too complex and still subject to research of proper mathematic model, you cannot expect the game to model it.
Besides, the difference is randomely distributed and overall insignificant to gameplay. At least to my gameplay.