Well the whole point of the study is that even if all the roads are gridlocked all the time, it's not as big a problem when the typical trip is shorter. If I spend 15 minutes in stop-and-go traffic and another 5 minutes in free-flowing traffic, that's better than someone who spends 15 minutes in that same stop-and-go traffic but then has another 25 minutes in free-flowing traffic to get where they're going. However, the Travel Time Index (TTI) would say the 40 minute drive is better because only 37.5% of it was spent in traffic, while the 20 minute drive spent 75% of the time in traffic. 40 minutes is still twice as long as 20 minutes, and could be three or four times as many miles.
So while there is still a lot of traffic in Chicago, the average commute is much shorter in distance than places like Charlotte or Houston, so the effect of that traffic on people is lessened. It's a case where intensity (traffic congestion) is tempered by the short duration (time spent driving in total). The ideal would be low intensity and a short duration, while the worst would be high intensity and a long duration. TTI is only measuring intensity, so it leaves out a big part of the equation.
More than anything else though, it does show that using TTI to compare different cities is flawed. Even when not comparing different cities, it can lead to some illogical decisions about highway building. You could have some DOT saying, "holy crap, the TTI is horrible! We need to build more capacity!" If the average travel time is only 5 minutes however, it's not worth the trouble.