This is a sample size of exactly two salads from one location. Which employee makes the salad probably has more to do with it than if it was delivery or in-person.
The full data set has 35 salads. The graph near the bottom shows 20 that were all for the same menu item. Of those 20, I'd say there were about 4-5 different employees doing the in-person orders. No idea how many different employees did the online orders.
Still not a _real_ study by any means, but there's more data than just 2 salads.
Yes, the lighthearted Medium post about food ordering is certainly the place to display your unwavering commitment to scientific rigor. We got a real intellectual heavyweight here everyone!
My experience is you usually pay more at least. Sometimes restaurants add a dollar to every item flat out on grubhub versus their actual takeout menu. Sometimes they charge another 4% for not paying in cash. Sometimes they do both of these things.
FWIW it's not necessarily the restaurant trying to upcharge you. The delivery apps will often take a percentage of the food costs in addition to their separate fees, so restaurants will have to raise the cost of each item by the same amount to make the same profits as an off-app togo order.
Berkeley and a few other cities capped the amount an app can take out of the restaurant's food costs, which is why you'll see extra fees when you order there. The delivery app is now taking their cut directly from the user instead of from the restaurant without the user knowing.
Still not a _real_ study by any means, but there's more data than just 2 salads.