We have so much information capacity now that's no longer an issue.
Gosplan, the Soviet state planning apparatus, ran on an monthly information cycle and a yearly planning cycle. WalMart runs on a daily information cycle and a weekly planning cycle. Amazon is near real time.
China, interestingly, goes to the opposite extreme. China is now on the 13th 5-year plan since the Revolution.[1] It's a set of 32 general policy goals to be emphasized for the next five years. These seem general, but this overall guidance determines on what the government spends time and resources.
[1] http://news.xinhuanet.com/english/photo/2015-11/04/c_1347835...
That's not the point. That information needs to be processed by a human being who then outputs decisions. If you have too much "big data", that data becomes useless because it overwhelms thereby making itself useless.
Only some of the information needs to be processed by a human being. How much inventory to carry of each item has been automated in retail for decades. Only unusual situations and new items need human attention.
Precisely this. Read Hayek's "The Pretense of Knowledge" for a more complete understanding of this idea: http://www.nobelprize.org/nobel_prizes/economic-sciences/lau...
Fundamentally, it´s an issue of scale. It can be proven that central planning is the least efficient form of organisation at scale, because the information capacity of a hierarchically organised network will always be less than that of a mesh or federal mesh system.
However for small systems, those network capacity limits aren´t reached - so it doesn´t matter, and there´s also a joker in the pack for mesh networks, in that if they can´t self-regulate their traffic, the resulting broadcast storm (which socially we are now in worldwide) will drive their good-put (user traffic) to very low levels.
Besides the information capacity issue though, there are also considerations of command and control, so if you have a central planner that really knows what they´re doing (and the issue of information capacity is essentially that most central planners can´t), then that topology will be advantageous. Especially in competition with other centrally planned organisations - like Microsoft say.