Here is some math for ya.
Distance on one battery: 18 minutes battery time at 20mph = 6 miles = 31,680 feet
Average distance for one delivery: 4,000 feet round trip max
Trips per battery: 31,680/4,000=8 trips before having to recharge at 30% battery
Okay, nice on the math, but that's presumably under ideal conditions with no worthwhile payload weight or drag factored in, nor wind, nor obligatory hovering or climbing to cruise altitude.
I'm skeptical that under real world operating conditions you would be able to do more trips than three safely, if that. Ten minutes per transaction (from power up, to return and power down) seems to be a reasonable time, so even under ideal conditions (just flying and none of the other concerns) you're at 30 minutes.
And again, what is being delivered? A pack of gummy bears? If you're talking about a lake or harbor, you're probably going to be delivering beverages of some sort and liquids are heavy - frozen (ice) - or not.
To make any type of service like this profitable, there's going to be quite the premium on the order. How many people are going to be willing to pay the price? We're not talking about captive audiences that don't have much of a choice (airports, entertainment venues, etc.) and can't plan ahead and bring their own supplies.
So I think:
1. With a
mavic 2, it is not physically realistic/possible
2. There's not a market for it. I don't see enough demand for what the
M2P can deliver even at a lake, etc., especially given how much a delivery would cost.
3. Running a service like this would make a lot of people unhappy. To make it worthwhile, you'd need to constantly have drones in the air. Depending on the situation and how dispersed people are will determine how much of a nuisance having these things buzzing around will be. At a lake, delivering to boats, maybe not a problem. At a beach, probably a problem.