Seems to me that many of the comments are failing to compare this model to the common alternative, which is a public facing restaurant. The failure rate of those is at least ~60%, so is this "cloud" model likely to be worse than that? I doubt it.

As a model, I think the customer here is not so much a food buyer but a restaurant starter and that is a bottomless well. This concept is a natural outgrowth of the current trend in quasi-legal "instagram restaurants" . It also meshes well with the fickle appetites and food trends of modern day. A cronut kitchen can sell out for 6 months and then fade into a ramen burger for another 6 months, all from the same physical kitchen.

I think that it is more sustainable than the food-truck -> sit-down-restaurant strategy that hasn't worked for more than a handful of entrepreneurs. American's love to eat a cornucopia of tastes and dishes but we absolutely hate to pay more than a nickel for it. The underlying cost structure of the cloud kitchen is more amenable to that sentiment.