I had the same reaction about the "closed loop" highlight in the title. An open loop system seems very wasteful and hard to maintain. Your comment on decentral

I had the same reaction about the "closed loop" highlight in the title. An open loop system seems very wasteful and hard to maintain. Your comment on decentralized application caught my attention. I agree with you. While there are clear merits to gaining economies of scale, there are often diseconomies of scale that go overlooked - which would make smaller decentralized applications more attractive. The trade largely comes down to lowering require capital expenditure ($/resource generated, e.g. $/kW) in order to gain operating flexibility (which can lower overall operating costs - though those savings are normally excluded from the published $/resource value, hence overlooked). Think of the merits of everyone having a car instead of trying to move people to-and-fro using trains. Sure, you pay less $/mile traveled, cumulative carbon emissions/mile are generally lower, and you avoid wear and tear on your car. BUT, you have to pay for taxis or rental car when you arrive at your destination, so you pay an inflated amout to restore flexibility. The cost for 'last mile' transit certainly isn't reflected in your train ticket cost. As a side study I've been wrestling with this problem in my industry (commercial ship design) for the past decade. One challenge is that it's hard for a corporation investing in the infrastructure to sell these decentralized units to achieve a profit margin competitive with a large-centralized product. The solution, I believe, is to figure out how to evalutate and factor in the cost of loss of flexibility in the $/resource calculation. That hasn't been an easy nut to crack.