Earlier quoted context omitted.
Sepp Holzer's farm is amazing. I think however it's unlikely that he's truly able to grow citrus fruit at scale even using all of his near magical techniques. Where he is able to grow citrus fruit is taking advantage of very small micro climates with ponds buffering temperatures and acting to reflect evening sun along with perfect orientation etc.
Permaculture in general sadly does not produce at scale at all and scaling up is very labour intensive.
Think about that for a second -- your garage is housing the equivalent of 120 horses. How much "labor" is that?
Now, we don't really notice that when everything is working well and the gasoline keeps flowing. There are calculations that figure a gallon of oil has something like two weeks of 8-hour/day manual labor embedded in it, energy-wise (using the cute term "energy slaves"). Synthentic fertilizer is basically spending a crap-ton of energy to fix nitrogen -- google "haber bosch energy".
Especially with the covid supply chain disruptions going on now, though, I've been rethinking what "scale" means or if it's even the right metric to be going for. Permaculture has been really interesting to me last few weeks because it doesn't try to go for "scale" per se (though there are claims that properly done it's more productive per acre), but rather it tries to cycle as many energy flows as possible.
So if your yardstick is "how many bushels of corn per acre can I grow?" then yeah, permaculture doesn't 'scale'. But if your yardstick is "how do I get the most output while minimizing energy inputs" and accept that outputs aren't necessarily a monocrop but can be complementary outputs (corn + legumes + animals + fertilizer), then it starts to get really interesting.
I'm a developer not a farmer, but the systems-level thinking I've seen in the permaculture community is extremely fascinating. There's 'labor' embedded everywhere, the question is are you counting it.