That was a *very* interesting article. The idea of a severely pruned citrus tree that is more like a ground cover than a tree would have particular potential in my climate where we don't get much freezing weather - just enough to kill and only on some years. Our problem is that when we do get snow, it's extremely heavy, wet snow usually preceded, followed, or mixed with rain. I don't think the suggestion of two layers of cheesecloth would do the job of "protecting". Every climate is different and has different challenges.
As the end of the article points out - the system required a lack of competition to be viable. Cheap food shipped from around the world would out-compete a system like this until the real cost of fossil fuels from cradle to grave is accounted for.
I saw that article a while ago, and it raised as many questions as answers. If you can get adequate solar inputs through the cold season, it could work. Big "if."
I wasn't clear on several things that seem pertinent, particularly latitude (meaning hours of overhead, high angle sun per year, and particularly during the freezing part of the season. Also, mean temperatures during the cold season, and expected maximum lows. The question is: can the trench capture and hold enough solarenergy to prevent a damaging freeze? Without becoming so warm that the tree breaks dormancy, or dries out too much? There are a lot of moving parts to this.
Still a cool idea, despite caveats.
EDIT: Reading it again, I wonder if some of the selected cold-hardy varieties still exist. The breeding program sounds interesting. I suppose one could also tease out climate data from the general description.
The only cure for that is hours of television radiation. And this tiny ad: