Earlier quoted context omitted.
b) 90% of Americans don't "do" traditional carpooling and they're not about to start 90% of Americans didn't pay non-professional strangers for a ride before Uber came along. 90% of Americans didn't share what they were having for lunch before Instagram. 90% of Americans didn't share the hilarious list of "cats who forgot how to cat" before Buzzfeed. The point being it's pretty much impossible to predict what product…
Your overall point is correct but I don't think it applies here. There are deep cultural reasons for why carpooling hasn't become popular in America. From what I have observed as a foreigner, people in this country love their big personal spaces, and their cars are a part of that. It's related to why public transportation isn't popular: most people don't want to sit within such close proximity to strangers.
There is a possibility that people wouldn't want to give up the convenience of commuting - having your car at the office so you can go somewhere on your lunch break, leaving the office without having to wait for your carpool ride, carrying large items to the office, and so on - are benefits that you would lose by switching to Lyft. People will have to weigh those benefits against the cost and stress of driving to work. So the question here is really "Do enough people want to give up driving themselves in favour of a cheaper, easier, but mildy inconvenient alternative in the form of Lyft?". The answer to that question is not an obvious "no". There could well be a big enough market to sustain a profitable business.
The fact that people haven't carpooled much before is not evidence that they don't want to or that they can't be persuaded to now; it is simply evidence that the pain of commuting was lower than the pain of organising a carpool. If you can lower the effort people have to put in enough so that it's sufficiently easy to carpool then it is possible change people's minds. Lyft will succeed if they can do that.