Fun run of thinking. I will argue that greenhouses are a future necessity. The world population is going to roughly double over the next 100 years. So unless we want to wreck a bunch of the natural world we need to produce nearly twice as much food on the same amount or on a lesser amount of ground. We can get 25% of it if we simply eliminate food waste. Where is the rest going to come from? Food forests etc will certainly be part of it. Their problem is labor and the embodied energy of that labor. Other things from other parts permaculture. More still may come Savory institute stuff on grazing. But in the end by the time we take more crop land out of service for homes and roads etc I don't think it will be enough. And increasing population means more shortage of water too. Greenhouses in norther climates should at least almost double productivity between being able to grow year round vs just 6 months. Now lack of light will limit that a bit but we get the gains of weather protection and the losses possible there as part of the picture too. The important thing will be optimal use of the acres under the green house.
Greenhouses are water and energy pigs if poorly and stupidly built. But that misses the point of all we have learned over the last hundreds of years. Greenhouses built and run properly actually use less water. So if we can capture the run off from the building we may actually be able to grow more crops or grow crops that are more water hungry without any increase in water needs. And by building them correctly their energy needs are small or they may even become net energy producers. And with proper siting they maybe be powered off energy sources we couldn't otherwise economically tap.
The one problem that doesn't have a permies compatible answer is glazing for large scale use. I will argue every other problem can mostly be solved with answers that are compatible or nearly so with permies principles.
We can start with simple answers like tree walls with no glazing. The low tech greenhouse that has been around since the time of the Romans. France has so many of them around some cities that the nazies blitzed around the town rather than taking on all of those walls. And other than the embodied energy to build and maintain those walls no more energy and because of micro climate affects already reduced water and rule of thumb says they are good for roughly one zone of climate improvement. No ongoing energy use and reduced water use. If we cap them with photovoltaic panels since walls that are durable long term need a roof with an over hang to help keep them dry they could be a net energy producer with no land lost to solar panels.
Now I will get back to glazing but will make one simple point. Glazing is also supposed to be good for roughly 1 zone per layer of glazing. If I run that up against a heat absorbing tree wall do I get 2 zones worth of gain?
So next you will point the structure of the green house using plastics or metals or wood that rots off rapidly not being with principles. Okay we can change that with many different options. For example black locust grows over much of north american and is extremely rot resistant. They talk about posts of it in swampy ground that are still good 50 and 100 years later. Surely is should be at least worth a look? If we choose a variety already bred for its straightness and height like the shipmaster black locust.(also does away with the problem of the tree spreading by seed) then we have another piece of the puzzle. We already know we can build laminated arches with spans in excess of 100 ft. That leaves needing a permies valid water proof, rot proof glue. We have options although here we may be better off mildly violating principles using modern glues for greater durability. Another option if we don't get in a hurry is we may be able to grow the trees to shape for smaller greenhouses
Next piece of this is end and side walls. Modern industrial green houses are doing them with tech answers. But here we can do a lot simply building with earth, rock, cob etc. And we can build machines to do them on a large scale with small labor. There again mostly in keeping with the principles. If we do it in a hole in the ground we make it easier to heat and make it so the resource to build the rest is already present. We are simply changing how it is arranged.
Now you say that still isn't good enough to be worth glazing the greenhouse. And I will say very possibly true. Next step is to add earthbanking, earth battery, GHT type systems. Most involve large quantities of plastic pipe buried in the ground. At which point you are screaming VIOLATION of permies principles. But then you look at the University of MN info on using rock beds made of of rock about the size of softballs to do the same thing. Sorting rocks of a certain size and then keeping them seperate from the soil does away with most of the problems of that one. Or alternately tapping local ground water for its heat. Either answer you will be hollaring is energy intensive. And I will point out that the power needs are comparitively small. If all that is running is a small pump or some small fans these are not energy pigs as expected. And if we can tap a local waste resource like micro hydro we might be able to source this much locally, or wind or photovoltaic solar are other poorer alternatives.
As for keeping the greenhouse warm adding an active foam insulation system using the dish foam type type system we add another active power element but being able to wrap green house in R12 to R30 or R40. This home has run all winter without any heat in northern WY. The coldest I have experienced this winter is 51 degrees. If we can put in say 5 to 10 times the sun by virtue of it being a greenhouse instead of just a home and insulate better at night in a green house it shouldn't be a serious freezing risk in any form no matter the weather.
And there are dozens of other systems we can add if we need more heat. Compost heat, RMH built into our earthen walls, Pain pile wood chip heat etc. Dealing with heating is one of the most important pieces of the puzzle but we have so many options. If the green house is going to run year round cooling is the next big piece of the puzzle. This is one of the places commercial greenhouses go wrong and become energy and water wasters. Using big fans to promote air change and water for swamp coolers for the incoming But if we design properly we can mostly eliminate those. The ability to open enough of it up for natural ventilation while still recovering as much outgoing humidity as possible can be designed for. Here we may still want those horrible earth tubes rather than the rock bed.
The point is that while this may not exactly be nature we can get benefits from it that I think far out weigh both the risks and the expense if we are smart about how we do it. The glazing is the only major piece of the puzzle that I see no semi permaculture valid options to get around.