Mathematical analysis of the problem leads to the same conclusion as I would have told if someone asked me about the food delivery business model in India.

The economic reason for possible failure can be stated in just a couple of sentences - Labour is cheap in India, so most restaurants keep on-staff delivery personnel. This makes it easy for them to ramp-up if needed and often they have no need to contract with a third-party, food delivery service.

I see this model as a possibility here in the US, where labour is not cheap, and not all restaurants deliver food themselves. And yes, I believe they can work well. I used DoorDash several times when I resided in an area they serviced. Whether these businesses flourish or not is an entirely different matter, and growing competition in the field is definitely not making it easy.