Aside the complexity and standardization needed for building it, which alone will absolutely not make it worth it, you'd also have the problem of temperature fl

Aside the complexity and standardization needed for building it, which alone will absolutely not make it worth it, you'd also have the problem of temperature fluctuations. You probably don't want your milk to be frozen right? ;-) So if you really need to push indoor air inside to avoid that, you are basically killing the efficiency. And not only that, you would push warm and comparibly humid air inside the fridge a lot. It will condense and now your fridge has to deal with a lot of condensation and probably get moldy. But good to see that I'm not the only one making these thought experiments. My personal conclusion is: eventually we will have a single refrigerant distribution system where other systems can be connected and then make use of it (similar to multi-split ACs, but extended to more devices). Then, your fridge would just be part of the system. And, just to understand: transferring heat from inside the fridge to the outside in the case that the outside is colder is so energy efficient already, there's no need at all to optimize anything. A solar panel as big as your hands will be sufficient even in winter to take care of keeping the fridge cool. That's how little energy is needed.