These days you can have an LLM code you up python to custom match visual metrics from your preferred web fonts to the likely user fonts across your statistical user base.
But why?
But why?