Actually, that graph analysis may be oversimplified. In September/October, it is HOT in Orlando. The lines for Splash are long even when the crowds are low because it is a wet ride and because it is very cool inside the ride, which lasts quite a while. In other words, people flock to this ride to escape the heat in those months. In January/February, it is cold. Crowds are higher, but the lines are just as short.
So, while there may be more marginally more people in the parks during that period, there could actually be fewer riders. But, more importantly, people will miss a cool/wet ride more in hot months than in cold months. So it will have a greater impact in the hot months than the cold months.
It isn't a simple math equation since the enjoyability of the ride varies based on the weather conditions.
This is, of course, just my interpretation of the facts. And, as
@rteetz suggests, this could be a part of a plan that I just fail to see or believe in. Because the only way the timing makes sense is if they simply cannot do this in January, when they have always done it, for some reason.