As an example, it looks like there are fewer than 300 zipcars in Chicago, certainly one of the two most well-served transit cities in the country. That might correspond to 5,000 users, so at best, about 1 in 1,000 of the people in the Metro area. Even assuming Zipcar users are 3 times as likely as others to use Amtrak, which I'm not convinced is true, you're talking about 1 in every 300 users of Amtrak who could even consider using a Zipcar at the other end of their trip.
That's in Chicago. There are only 50 zipcar cities nationwide. Bloomington, IN, has only 4 cars. Elon, NC, home of Elon University, has 3. (I guess it's also possible you thought there was a focus on academic types because of the number of college towns listed, but actually quite the opposite. The focus of these cars isn't liberal academics. It's middle-class college kids who otherwise wouldn't drive at all while they were at school. And the zipcars in these places are more like recruiting tools than profit centers. The idea is to plant the seed while kids are in college.) At any rate, the point being that there are precious few actual Zipcar members, so a Zipcar strategy for Amtrak makes no sense. It can't help enough of their riders to matter.
Zipcar may find it profitable to have an Amtrak strategy, though. That, I think, is what is playing out in the posts above. Having a Zipcar near a handful of Amtrak stations certainly doesn't hurt Amtrak, may help a tiny bit, and it helps Zipcar a lot, because it helps Zipcar members feel like it can really serve nearly ALL their car needs.
We were actually thinking of joining Zipcar, or the local competitor I-Go. But then my fiancee's parents gave her a hand-me-down car. We wouldn't have bought a second vehicle (I do have a vehicle - a pick-up I need for my business, but I drive it very little outside the season - I didn't even start it in January.) but having been given one, it's more economical to keep it than not.