You're assuming that Fastpass accounts for half of the capacity of the ride. Its actually closer to 70, 80, even 90% when things get really backed up. So if a high demand ride has a 1,500 hourly capacity, somewhere around 1,125 (75%) of the riders will have come from Fastpass.
You can't look at it as just a slice of one hour. Let's say the system distributes 1,125 Fastpass for 3:00 to 4:00. In your explanation, only those Fastpass holders will enter the Fastpass queue during that hour. But what about 3:30 - 4:30? 3:35 - 4:35? etc. The result is people
don't return in an even flow of equal to 75% of the ride's capacity for that hour, they return in a higher number that would ever
physically be in front of you in the standby queue. To keep their Fastpass wait time to a minimum, the merge point ratio is increased to accommodate them. So now the standby line is moving slower than it would if
only the 1,125 guests were returning. So it doesn't "even out", it becomes artificially inflated.
- the standby queues are not 50-75% physically full, they're generally the same physical length they always were. Only now with artificial inflation. So it's more like this:
View attachment 57501
Besides, the data is right in the link in the OP. The major attractions have less Fastpasses entering the Fastpass queue and are seeing a reduction in standby wait times, meanwhile rides that didn't previously have Fastpass are seeing an increase. It isn't because less people are also using the
standby queue, because, like in the case of Soarin'/Test Track and only being able to have a Fastpass for one of them,
more people are likely entering the standby queue than before.