This seems like a huge impact. Not sure how they came up with that number (I'm guessing customer surveys, so it's probably not that accurate).
How does this change the debate that ride-sharing congests up cities?
The only possible benefit is if people take public transit more or take fewer trips, which is not something you can see from ride sharing numbers.
It only makes sense if they are transporting two or more people that would otherwise have used two or more vehicles. But I think that kind of rideshare system is the absolute minority of Uber and Lyft rides.
And if you ditch your car in favour of ride sharing, then you're likely to make less trips overall. You'll think twice before driving somewhere you could otherwise walk, more so than if you owned a car. And you're less likely to make impulse trips that can be avoided (to get a chocolate bar you're craving or something else pointless).
At least, I'd imagine that's the case - as there is the additional time of hailing a ride with one of these services and a more measurable cost to it.
You forgot "circling for twenty minutes searching for parking." Include the parking footprint of private cars and I think ridesharing comes out ahead.
The general case is driving to a parking lot/garage which could make the trip shorter or longer depending on where it's located relative to your destination. You might spend more time driving in a garage, but that does not impact street level congestion compared to all those extra Miles drivers add going to pick you up.
As to density, cars parked in a garage take up vastly less land than than moving cars on a road as they stack vertically and need less space around each other.
A) from less usage because of Uber
B) because all these extra taxis have caused mass congestion which means buses are unbelievably slow compared to even a few years ago