There have been a lot of answers that boil down to:
- Just do it for the $$$
- Do a mock up of something better
- Run a usability test
- Run like hell (but politely)
All of these assume that:
- The design really is bad
- You are objectivly approaching this project
Please don't hear what I'm not saying: I am not saying your premise is wrong.
First, I am saying its worth considering that it may not be that bad. There are a lot of very successful sites out there that drive designers nuts. If it doesn't satisfy a need, it won't matter how well its done.
If you want to be a part of this than you need to look at the business plan first, and see if its valid, and if the site will fulfill it. If either one of those is a no-go, than fix those problems first.
Second, ask your associate if (s)he's considered other designs. Ask for the sites they used to come up with what they want, and ask what else they've considered.
If they give you something, look at what makes those good/bad and see if between the two of you you can come to a better design.
If they can't, then tell them that with some work there may be something better and work together to create 3 or 4 mock ups. Then whip out the usabilty test. Even if you've done them before, I highly recommend reading what Jakob Nielsen has to say about it over at http://www.useit.com/alertbox/
Since this seems to be a business partner, just doing it for the money and/or running like hell doesn't seem to be an option. Working through this with a little more give and take may help, but if you can't come to an agreement, then you probably ought to leave.