Interesting that you took a task in that range.
I remember having read somewhere that peoples ability to estimate something declines with increasing size of the estimated thing. So, applying this logic, the task you estimated would be one where relatively accurate estimation should be possible.
Now, you actually did face the effect of uncertainty. And of course the uncertainty for individual tasks has effect on estimating a project size. After all, that’s the crux with uncertainty: you don’t know when it strikes. It can happen with all or none of the tasks. It can add a few minutes or even weeks.
But the problem is not estimation itself. It is using estimates for planning.
An estimation, even if accurate, is basically describing the ideal time to completion under ideal conditions. Executing the work doesn’t happen under ideal conditions.
Like, if you estimate a task with 8 hours it would take 8 hours to complete, if you could work on it 8 consecutive hours. Focused. Without interruptions. Without any surprises.
If you try to plan on estimates you need to account for each of these factors. And that is the hard thing. People tend to account for that with static buffers, but that tends not to be sufficient in reality.