Second, yes, by definition regenerative agriculture is scalable.
It might require as much as double the labor requirements (so 3% of the population rather than 1.5%) but it's likely that automation will amortize that.
See "How is China able to provide enough food to feed over 1B people?" https://news.ycombinator.com/item?id=20537409 which describes some aspects of China's agriculture that are regenerative.
Also, I recommend Toby Hemenway's "Redesigning Civilization with Permaculture" talk for background and future directions. http://tobyhemenway.com/videos/redesigning-civilization-with...
Geoff Lawton's "Greening the Desert" project converted salty desert to producing figs and other food in just two years. https://permaculturenews.org/2007/03/01/greening-the-desert-...
- - - -
Just to mention one (gross) example: It's a little-known fact that Eisenia fetida worms will devour "night soil" (#2's) and produce clean-smelling castings, while doubling in biomass in approximately 40-50 days. This could easily be scaled to handle urban sewage densities, returning that matter directly to the ecosystem without expensive treatment/chemicals etc.