I saw some examples of this while travelling in Japan last year, and am keen to replicate. I suspect, like a lot of things, the local product there is superior to what I could find back home here in AU.

At some castle courtyard, we were walking around on a smooth-white-gravel surface, and I didn't immediately register how different it felt.

Stare at the ground for a few moments, and notice some of the small (15-20mm) off-white pebbles are perfectly aligned on a grid pattern.

Squat down to investigate, and they've got these corners of the plastic mesh geogrid topped off with little white caps that are almost indiscernible from the rounded white aggregate they're using as infill. It was almost like seeing a random dot stereogram, except completely regular (and no dinosaurs) of course.

The math is pretty interesting, and TFA didn't touch on this aspect I don't think. I'm looking to use 10-20mm irregular shaped scoria on a large-ish area, and the heuristic evidently is that your maximum aggregate size should not exceed one third of the geocell depth - so I'd want 50mm depth at most there, sitting on solid-ish sub-layer.

(I believe that heuristic goes out the window if you've got regular shaped gravel.)

The other ratio to consider is aggregate size to cell size, and that, in turn, is dependent on load and slope. Then there's the mechanical attributes of your in-fill (will it slowly crush over time, with expected loads - for softer rocks you'll definitely need a geotextile above your sub-base and below your geo-cell.) All tremendously more complicated than the 'few wheelbarrows of blue metal' I grew up with.