A few people wrote a paper outlining a decentralized model for economic planning based within the "communization" tendency, wondering what people will make of it:
https://journals.sagepub.com/doi/full/10.1177/10245294231213141
It is interesting to see cyber-physical systems being invoked explicitly. I've been working on a text that does this and connects it to planning. The electricity grid and how it connects to industry is particularly relevant. I see that they point out that I-EPOS could determine appropriate schedules for running loads. Which is an interesting proposal. But, horror of horrors, that would be centralized!
In reality there is no contradiction between fully centralized accounting and computation (logical centralization) on one hand, and decentralized politics on the other. In fact, full logical centralization can be shown to be mandatory (Conant-Ashby - Every good regulator of a system must be a model of that system).
This brings me to a point I've been wanting to make for a while: so far most proposals I've seen of "decentralized planning" have one thing in common - being Worse Than Gosplan. Theorizers of such decentralized planning fail to understand how the Soviet system actually worked, that it was neither centralized nor ex-ante planned. The charge of centralization applies neither in decision making nor logically. The system was very much dispersed with plenty of horizontal coordination. In short, a viable system. It did have some obvious issues however, of which two are: long delays and asymmetric access to information (data bunkers/tiny popes). Luckily this paper doesn't suffer from this nonsense quite as bad as other texts I've seen.
The paper seems to conflate supply/demand and remuneration. But in the end they seem to come to the correct conclusion: that there need not be any such conflation. Which is true of course, these things are orthogonal in planning.
I see the paper references Sutterlütti and Meretz, whose work seems to boil down to much the same as a lot of decentralized thinkers: replacing computation with endless meetings. The thought experiment in the paper seems to be one example of this, and will inevitably run afoul with Dunbar's number. These thinkers never address how products with long supply chains, which may involve millions of distinct intermediate products, are to remain feasible. You will not be able to find someone willing to make U235 for your fission reactor by just asking around.
TL;DR: interesting paper, makes the case for logical centralization without realizing it.