I have never liked the station wagon analogy, because it misunderstands the thing we are trying to measure. In the example, we measure the bandwidth of the station wagon. But that's like measuring the bandwidth of a packet -- a nonsense concept. We measure the bandwidth of the channel, not the chunks of data which fly through it. To really get the right analogy, we should talk about the bandwidth of a freeway, not the station wagon which drives upon the freeway.
Bandwidth in the colloquial sense means "the amount of data which passes a given point, per second." So, imagine that you can load 25 TB in the form of tapes into a station wagon. For safety, these station wagons must drive a distance of 75 meters apart and a speed of 100 kilometers per hour. That means that one station wagon passes a given point every 2.7 seconds. That's 9.2 TB per second. Adding a second lane to the highway would double the bandwidth.
The stupid calculation which is often performed, on the other hand goes like this. You have 25 TB in the wagon, and you drive it to a location 10 hours away... Already you've gone off the tracks, because you are mentioning the TIME it takes to get to the destination, i.e. the LATENCY. And as anybody knows, the latency (or equivalently the distance between the points) has NOTHING to do with bandwidth.
How can you say Time has nothing to do with bandwidth when, in your own example, you measured it in TB per SECOND?
Following your example again of 9.2TB/sec, that can be changed to 9.2TB * 60 /min, or 9.2TB * 60 * 60 /hour, or 9.2TB * 60 * 60 * 10 / 10 hours, which is the exact measurement that you seem to have a problem with earlier in your post (data in a 10 hour period).