This looks like the poster child of a solution looking for a problem. It's all based on Adobe's development of Multiple Master fonts in the 1990s, and if it hasn't taken off in that amount of time I doubt it ever will.
The use of web fonts has never been greater. With more websites using non-system fonts, Variable Fonts serves to reduce page weight while providing a technical and design solution.
If a website uses more than two weights from a given font, then a Variable font would save bytes downloaded. The desire to optimize websites is not going away soon, so this will continue to be valuable.
The unnecessary design restrictions around “number of font weights” that many web developers are used to would be blown away by well designed Variable Fonts. Too often web developers are tasked with limiting a designer from using more than two fonts or choosing between a limited number of font weights.
A single font file can replace multiple fonts by utilizing the expanded design space.
Additionally, when considering non-Latin fonts, there a huge implications of using custom fonts. Font weight improvements there are in the megabytes.
Beyond the utility of font size, Variable Fonts present the concept of controlling text within a multi-dimensional design-space, rather than a pre-designated font instance. This caters new opportunities to improve readability on various display sizes.
With the growing number of contexts where people consume text (small and big screens/AR/VR/IOT) the added control provides new creative opportunity and routes for accessibility.
The biggest holdup on variable fonts is that they're simply a lot harder to design than regular fonts. The lack of usable examples will hold back the technology more than anything else.