Unquantifiable and Project Management

Today I have been thinking about one of the dilemmas most projects face: quantifying the work and estimation. Even an experienced project manager or team leader can never be sure to make the estimations correctly. Reasons for that are many: there might be different people each time, or different teams within the project, there might be different requirements, different business context, different technologies and of course different priorities, responsibilities, goals.
Recently, at my work we started a new project and this is how we made our estimations:

1. We tried to size the project relatively. This new project that we took up to realise was a remake of the old companies website. Having the information about how long that first project (old website) took, how many people were working and how many we have now we could estimate how long the new project would take. It's hard to estimate without having some kind of comparison point, thus estimating comparing to some similar project can help to quantify the new project.
2. Being open and honest with our stakeholders. At my work, we have an agile working philosophy which fundamentally is based on transparency and visibility. We openly discussed with our stakeholders that we are not sure how long this project will take, we can only give our best guess. However, with some iterations we can build something and then measure how long it takes, which would eventually give a better sense for estimations of the entire project.

In other words, we used cone of uncertainty for making the quantified estimations for our project.

"This phenomenon is known as the “Cone of Uncertainty” which is illustrated in the following figure. As the figure suggests, significant narrowing of the Cone occur during the first 20-30% of the total calendar time for the project" [1].

The Cone of Uncertainty suggest not to settle down the deadline for the project upfront but rather iterate, create prototypes and elaborate before even starting to construct detailed plan for the project.
I am very curious to hear from you, my fellow colleagues on how you, or your teams deal with the unquantifiable. Anyone has some examples? Would like to hear!
Resources:

1. https://www.construx.com/books/the-cone-of-uncertainty/