Problem with undergraduate research supervisor, how to proceed?

I'll be honest with you, remarking that this is only my opinion, by no means I'm right.

Undergrads are a pain to supervise.

Mostly because, in general, undergrads require a lot of "hand-holding" and training (proper research procedure and writing are not usually covered before that), which takes time (even worse if it is a good student that challenges you.... rewarding, but takes even more time).

You are right in thinking that undergrad research should be simple, self-contained, and well planned out. Properly planning stuff takes time, professors never have it. And simple means that the research is unlikely to have high impact.

I know several professors that don't accept undergrads simply because they think it is not worth the effort. The ones that do accept, pin them to a Ph.D. student (or a postdoc) and that's it. The Ph.D. students usually are busy, worried about their own things, and do not have the experience to deal with "underlings".

Yes, your project should be well defined, with clear goals and schedule. That never happens. Not as an undergrad, MS, Ph.D., postdoc, or professor. There might be a few unicorns around, but, usually, you have no idea where you are going. Ideally, you could use this opportunity to train the most important trait in a researcher: independence.

Instead of expecting to be handed goals and schedule, create them. It will be challenging for you, but you would learn how to do it, and get familiar with the field. Then you could go to the Ph.D. student and ask "Do you think this is a reasonable plan?" instead of "what should I do?". Worst case scenario, that plan can be a starting point for a better one. Bonus points if you include solutions to tangential problems of her research (stuff that would be cool/easy to do, but isn't a priority, so she would never do it herself).

Quoting a far better man, and professor, than myself:

"Be honest, be kind, be useful, be responsible, work hard, treat everybody with respect."

I'd advise you to go around the Ph.D. student only as last resort.

You correctly identified (i) that the solution is to communicate, and (ii) that the problem is that the grad student might be inexperienced in advising students, though probably qualified as far as the material is concerned. You've also correctly understood that research is often an endeavor that takes you where the data and insight takes you, and that that might be a different place than originally planned.

You're both adults, and in a sense colleagues because you're still starting researchers. Ask her whether it would be possible to have coffee to talk about your project. Get away from the lab so that both sides set aside 30 or 60 minutes without feeling that every minute spent on talking is a loss of time for something else. Brainstorm about what you're passionate about, what she thinks this could turn into regarding a project, and jointly come up with a timeline. If you want, have a bullet point list of things with you that you think you need to talk about.