I found this thread interesting and decided to revisit with newer, objective technology. AI Face Symmetry Analysis and Proportionate Interpolation coupled with Face/Emotion/Ethnicity Recognition. From this the list of suggested most beautiful can be ranked at how well the proportionately fit the ideal universal constants for beauty.
Just for kicks I also placed typical "most beautiful women samples" into the group as checks and controls. These are women who have been agreed as being very or the most beautiful repeatedly.
In every case I used a range of images, between 3 and 5 from a range of ages and poses and took the average. For some women, who had much longer or variable careers I did include samples from their high points of exposure. Surprisingly, only a few changed score over the years. Even those whom I sampled across many years only shifted a point or two. I included the exceptions in this graphic just for discussion.
The score is the average of the images, and the number is the degree to which they fit a mesh fit matrix of dozens of control points to their own faces. This is not how well they match to some mathematical golden ratio like in years past, but rather is matched to the perspective of the image. This is much newer analysis and is surprisingly accurate.
The review looks are thousands of relationships and determines how much the face diverges from symmetric as well as the human averages for the identified heritage, age, gender, and emotion being expressed. So, there is much less of the, "Oh their head is turned or hair is in the way or she is smiling".
I will mention that one model did hit as "male" on a review of one of their images - I wonder if you can guess which one? The rest of her images hit as "female".
Liza Soberano is on the list as she was ranked as the worlds most beautiful woman in 2018.
A score of 100 is perfect. Nobody scores this unless you are a computer generated image.
The answer to this thread is: Lorna Morgan
In my review this total fits what the AI was telling me.