Look at their business model (at least for SpoonRocket). They're directly addressing the production cost of food - limited menu (two choices a day), menu driven

Look at their business model (at least for SpoonRocket). They're directly addressing the production cost of food - limited menu (two choices a day), menu driven by availability (seasonalize ingredient cost), highly automated delivery (meals loaded in car before sale - use the car as the warming bin). Limiting to two meal choices a day isn't an option for most restaurants. This is a significant efficiency optimization in manufacturing that also drives significant optimization in delivery. This doesn't strike me as a dumb hack at all. Quite the contrary, I'm very impressed. I wouldn't be surprised at all if they're actually profitable from a production/delivery standpoint, and the losses are just software and market development. (I have many years of restaurant experience in addition to software experience. I see the optimizations. Wow!)