How nonsensical to have this as a percentage. The underlying operational cost doesn't really depend on how much money you spent on the food. I don't see how delivery can be profitable on < $10 orders with a 15% cap.
I ran single digit labor in our restaurants, and this wasn't negatively affected by the number of deliveries because drivers contribute labor elsewhere.
If you think it's more cost effective to use a delivery service in lieu of staffing, then you'll need to compare against the cost of not having that staff and the sales volume increase having delivery provides.
During the pandemic, fewer family owned restaurants shut down than corporate chains in my county. When talking with owners, each said that delivery services had at minimum doubled their average dine in revenue despite fees. It was less expensive than employing the people required to run deliveries at their volume, and less costly than if they had attempted to do it with existing staff.
You don't want to fire people in a recession or a pandemic, and this was a way that many of those businesses could keep everyone on the schedule without taking on the risk and frustration of attempting to hiring new people who might not work out (or who might have car problems on the same day) or good people who they might not have enough hours to schedule.
Also, some fees are flat, and the percentages are there to compensate. There are also minimum pre-tax totals to qualify for delivery starting around $20-30, so your scenario of $10 delivery orders doesn't apply.