Good explanation here, and it’s the posted wait times that I was thinking about when considering manipulation of wait times.
As to Disney trying to argue their poor estimation of wait times is because of legitimate reasons, then they just need to actually measure the wait times! It shouldn’t be hard. In fact I’m sure they do measure them, and then inflate them. Either that or they’re incredibly bad at doing what Touring Plans do very well.
The problem is more complex then that. This is a good example of "The most precise answer is not always the best answer".
Note what Touring Plans does is very different. Touring plans makes plans with PREDICTED wait time based on historical models and data input, and constantly measuring things based on crowd sourced data. They also report current wait times based on crowd sourcing and prior data. They basically are posting predictions based on modeling and refine that model all the time with crowd sourced (and measured) data.
Customers expect Disney to be showing the 'current wait'. The problem is any wait you measure is actually in the past - it is not actually what a new person getting in line will experience. So the most precise answer, may not actually be a good indicator of future waits. Posting wait times is always an estimation game because you are not explicitly counting people getting in line, especially with extended queues.
There are legitimate reasons for over-estimating - like
2) customer psychology. Customer sentiment will be better if waits are lower than expected then they will be if waits are under estimated
3) over-estimating reduces probability of missing targets due to variations and disappointing guests
4) Just like TP, Disney maybe building anticipated factors into the wait time. 5mins after the park opens, the line is growing far faster than any count makes reasonable. Should the wait board say 5mins, then keep climbing every 30 seconds? Or does it make more sense to estimate the line will quickly get to 60mins and post that knowing that will happen.. and some people wait less.
5) Just like TP, Disney can know through modeling what the demand will look like for an attraction and build that into the prediction, rather than waiting for the result.
If you just relied on measured waits, your results can be an hour or two stale!
Instead estimating wait times is more about reliable prediction.. with limited measured points. And the penalties for under estimating are far greater than for estimating too long.