Did you make manual MPG calculations for both cars? If not, why should we think the measurements are accurate at all?
Yes, I did. The Jeep was relatively "on", within 0.5mpg typically, while my CX-5 is a bit pessimistic by 1mpg.
To substantiate a claim like this you'd need more accurate data, better not just one vehicle (e.g. maybe yours is broken?). Sorry, but 1 sample is just meaningless and relying on inaccurate dashboard gauge for this makes all this just pointless.
"Maybe my CX-5 is broken". Well, that's pretty bad, huh? It's "broken" even though no SES lights are on, it runs smoothly, etc. Either you're very wrong, or Mazda has magic gremlins.
Also, it is ridiculous to compare a gas-guzzler, with no efficiency what so ever, to a CX-5, on fuel economy when speeding. It is not meaningful at all.
I'd not consider my Jeep a gas guzzler. I got around 16mpg out of it during my daily commute while the CX-5 does 22.5. That's not really offputting to me.
Compare the MPG distribution of a direct competitor with similar configuration, the
CR-V vs. the
CX-5.
What you notice is that the CX-5's fuel-ups bars close to the average are higher then their neighbors, where as the CR-V gets higher bars away from the average, by a little bit (compare 22MPG bar at the CR-V is almost 1/2 that of the average, same bar for the CX-5 is clearly less than 1/2 the average). This implies a higher variance for the CR-V. Now, I am sure some people on fuelly drive faster than 75MPH on a daily basis, for both CR-V and CX-5 groups, only that the CX-5 distribution is somewhat tighter around the average.
Who knows? It's hard to say, because there isn't a "highway only" comparison.
If the CX-5 was finicky about getting best MPG in only very narrow set of conditions, the distribution would have shown it. It does not.
This is based on 128 different vehicles and 7470 fuel-ups for the CX-5, not 1 sample.