A)
If a beverage has a certain number of calories from sugar per volume, it doesn't really matter whether it's cola-flavored, orange-flavored, vanilla-flavored, tea-flavored, natural-flavored, artificially-flavored, carbonated, uncarbonated, caffeinated, or uncaffeinated. They're all functionally soft drinks, from a dietary perspective, but the graph that charts the decline in carbonated soda does not show the other members in this group.
.
B)
I actually harbor a hypothesis that I have not seen explored which would explain this even if we take it on face value. Contrary to traditional sources of sugar and starch, sweet beverages cause an absorption of sugar unimpeded by the normal vagaries of digestion - there's no chewing, nothing to slow a person down, no cell walls to slowly break down in the stomach, no fiber to absorb and slowly squeeze out the calories.
In the 90's, I used to drink two or three liters of birch beer for something to do while waiting for the pizza. You can keep pouring orange soda into a human body without complaint, long after they would have vomitted up a sugar-equivalent cherry pie. And often our behavior shifts to adopting soda whenever we look for a drink, as the default, several times a day.
We know that blood sugar spikes cause insulin spikes, and we know that ghrelin, insulin, & related hormones act together to modulate the hunger & satiety responses.
My hypothesis is that rapid sugar intake via drinks, when combined with our natural impulse to slake our thirst, acts to dysregulate our hunger and satiety responses in the longer term. Through whatever processes, it detaches hunger from ad libitum consumption levels, sufficient that an extra few hundred calories per day is the 'natural' level that affected people, on average, feel they should be eating. 100 extra calories a day over burn rate is about 10lbs/year of gained weight.
.
C) Obesity is a lagging indicator here. People faced with an ad libitum diet find it very, very easy to maintain their weight. There seems to be a nearly universal misconception that baseline gluttony is associated with high weight levels, but that's bullshit. A person weighing 300lbs has roughly the same neurochemical hunger/stress response as a person weighing 150lbs, when you shortchange them each by 500 calories a day relative to their metabolic burn rate: It's just that metabolic burn rate goes way up as one gains weight. Maintaining weight is the default state. The mystery is why people are on average slowly gaining weight, rather than why fat people are not losing weight.