One, the speed vs economy curve is different for every vehicle. That particular vehicle does best at 55mph and 25% worse at 75, but that doesn't mean every vehicle will.
Two, that graph is a textbook example of how to lie with statistics. The range begins at 15mpg, so a roughly 25% drop in real numbers (from 30 to 23 mpg) is represented by a roughly 50% drop on the graph (from 15 units above the X-axis to 8 units.)
Three, a 100 mile trip in this hypothetical car at 55mph and 30mpg takes 1 hour and 49 minutes and burns 3.3 gallons of gas. At 75mph and 23mpg the same trip takes 1 hour and 19 minutes and burns 4.3 gallons of gas. $2.50 of gas to save half a hour sitting in a cramped car is a good trade off to me.
Three, a 100 mile trip in this hypothetical car at 55mph and 30mpg takes 1 hour and 49 minutes and burns 3.3 gallons of gas. At 75mph and 23mpg the same trip takes 1 hour and 19 minutes and burns 4.3 gallons of gas. $2.50 of gas to save half a hour sitting in a cramped car is a good trade off to me.
Another good point.
Of course for those working near minimum wage, "earning" $5.00/hour in gas savings isn't that bad. So it probably is worth to to them drive 55mph instead of 75mph to work.
But I'd also be interested in seeing what the increase in demand for gas does to the total price for gas. If everyone were to drive 55mph instead of 75mph, consumption would drop and the reduced demand for gas would lower the overall price per gallon so the savings would be greater.
Of course a single individual driving at 55mph would have little to no effect on prices, but a whole nation slowing down and reducing demand would.