Your glass, sir, is half empty.
A window of an hour is easily in conflict with any number of activities during a day's park visit.
A window of 2 hours is in conflict with much less.
No - simply the idea that extending the window 'solves the problem' is not true. Your idea is premised that the problem is purely based on conflicts with fixed things - when in reality a big portion if it is related to the mentality that people know they have a big window and have the tendancy to relie on the window's length, rather than return at the start of the window. If you just make the window bigger, all it means is people will procrastinate more.
The problem is people say 'what time do I need to be there BY' instead of 'what time is my appointment'. People look at the LATEST they can return.
It may minimize the problem for some, but only for those who make the conscious decisions to prioritize the ride in the first place. By making it a bigger window, it simply reenforces the idea that the commodity isn't as limited, so people will treat it with even less importance. 'Oh don't worry, we have 3hrs to use it!' and they'll still make the same choices, like leaving the park, that will cause conflict with returning within the window. 'We have three hours, no need to goto the ride NOW, we'll do it later', etc etc etc. It will just be more reason for people to procrastinate. So instead of solving the problem, you've just altered it and the lingering problems still exist. It's a hack job that doesn't make the problem go away.
The way to solve the problem is to allow people to defer to a later time window. The key is how to do this without requiring a huge investment in people and resources and not slow down the vast majority of transactions.
Here's the scenario... the current return time is during your upcoming ADR. You get a FP ticket for the current return time. Imagine if the system would allow you to insert your current FP, and exchange it for a FP that is automatically 90mins later. As long as there is a later window available, it allows you to exchange your current allotment for a later one.
This solves problems with your 'I get there and I can't use the FP because the return time conflicts with my plans' scenario. It also solves problems like 'our plans have changed, and I can't ride right now...'.
This requires no user input and no CM intervention. All you need is FP machine that can dispose of inserted media instead of always returning it.
If you wanted to extend the feature to include guest recovery... you allow people to exchange a FP for that DAY, but may be expired, for a future time window (the same 90mins from now, or possibly if exchanging an expired FP, you give them a FP for the next available window. Customer Sat decides if you allow a override FP if no more time windows are available).
Such a model gives the guest what they want (a valid FP for a different time) without ignoring the idea of fixed capacity per time window. Making the window longer screws with the predicted # of people returning per window... simply giving you a different window does not.