Although I am all, "Whoo!" about finishing the layout, coming to the end of a project actually dredges up all kinds of anxiety for me. Part of it is just anxiety over actually putting the book out there To Be Judged. Part of it I think is a relic from my freelancing days--as long as I was working, I knew I was getting paid, but when a project ended, I had no guarantee that another one would come along. Part of it is workaholism--when this is over, whatever will I do with my days? (Hint: Start the next book!)
I know I'm not the only one who sees the end of a project coming and goes, AIIIIIGGGHHH! Matt Groening has a great cartoon on being a graduate student that says in part, "The Simple Way To Avoid the Stomach-Churning Agony of Having To Finish Your Thesis: Read Another Book. Repeat When Necessary." The Perfectionism-Procrastination-Paralysis trifecta rears its ugly head about now. Especially if you're unaware of what's going on, it's really easy to get caught up in getting just one more read from that cool person you met in writers' group...and then just one more read from your old professor...and then just one more read from your Aunt Edna...and then just one more read from that guy behind the counter at the gas station who seems pretty articulate.
You just have to soldier through and GET THE BOOK OUT. It's especially tough when you aren't in a traditional work environment with a boss and explicit deadlines--even (or especially) your friends will enable you by saying things like, "Hey, you sound unsure about this. Do you want me to read it over again? I'm pretty busy, but I'll have some time off next Christmas, so I could read it then." Beat them away with love.
There is no such thing as a perfect (or even just an empirically good) book, so let that go. What traditional publishers do is they have a process. For example, at my first job, the manuscript was proofread three times, and then it was laid out, and then it was proofread two more times. And there were still errors in the final book, but it was pretty darned clean. So I would suggest that if you're having problems with eternal re-reads, you should set some limits--you'll have it read X number of times by X many beta readers, and then it goes to a proofreader. And that's that. You set up a process you can trust, and then you do indeed trust it to get you a not-perfect-but-reasonably-good end result.