I wrote this elsewhere, but there's actually a lot else at play. And if you're female, you'll get extra pressure from family, friends, significant other, etc., to choose the school that is close to them. Essentially, if you opted for a school in the 10-20 range for personal reasons (e.g., family, proximity to significant other, other obligations), you've opted into playing on super-hard mode, where you have to work twice as hard and be twice as good to get as far as someone who opted for the best school to which they were accepted.

I definitely don't think I would have appreciated the difference if I hadn't experienced both in such a short time frame.

The top universities often:

1. Send grad students to conferences (including travel costs) whether or not they have results to present that year. (If you're at a slightly suboptimal school, you might only get to go if you have results to present, which you probably don't your first year. My labmates and I would volunteer and spent a day and a half stuffing swag bags and registering attendees/handing out badges for a conference so that we could get access. Students from better schools just had their tickets covered by their schools.)

2. Hire lab techs and admins, giving grad students more time to do actual research or write papers. (Otherwise, you might spend more than 40% of your time doing work that doesn't move the needle for you, professionally. You may be picking up your professor's dry cleaning, or taking care of administrivia around getting them a visa for their next talk in another country, or booking them plane tickets.)

3. The name of the school makes it more likely that the student gets a grant proposal accepted by the NIH/NSF/other agency, even if the quality of the grant proposal itself is the same. Getting a grant/fellowship/award on your CV makes a HUGE difference when applying to post-docs or faculty positions. (You'll also have access to more people who will give you more valuable feedback for how to write a better grant proposal.)

4. Top schools often protect students from bureaucracy and streamline the procedures in place for everything from getting access to a new facility, to what happens when you need to replace your student ID. (See (2) above.)

5. You're more likely to get a smarter or harder working undergraduate student to mentor and help out with your research. This matters more in the life sciences and fields with "benchwork," (e.g., wielding pipettes) but it applies to tasks that involve coding as well. In better schools, a larger percentage of undergraduates opt to get some experience in a research group, and you get both someone to delegate actual research tasks to, and you get to level-up in people-management.

There's a lot more, but you get the idea, I think.