A single mpg figure rarely covers a whole trip - city stretches drink more fuel than highway ones, and fuel prices shift as you cross state or country lines. Break the trip into legs and get one real total.
Highway mpg is typically 20-30% better than city mpg for a conventional engine, since steady cruising keeps the engine in its efficient range while stop-and-go city driving doesn't. If your route is genuinely mixed - a city stretch to get out of town, a long highway run, then a hilly or mountainous section - using one blended number for the whole trip either overstates or understates what you'll actually spend. Splitting it into legs and using the efficiency you'd actually expect for each one gets you a real number instead of a rough guess.
The same logic applies to fuel price on a longer trip - it's genuinely common to cross a state or provincial line partway through and see a real difference at the pump. Giving each leg its own price captures that instead of averaging it away.
The split at the bottom is deliberately a simple equal share, not weighted by who's getting dropped off first - that's genuinely how most carpools actually divide gas money, and building in more complexity here would be solving a problem this specific moment doesn't usually have.