Peer review DOES lead to better code. At college, my classmates and I would exchange code frequently, we would criticise moderately one another's code, and suggest improvements. Sometimes the suggestions were taken into account by the original author, sometimes by the proponent.
In the end, our applications were more featured and robust than any of us would've done by himself, each of us had a different point of view and could see different problems that could arise, as well as propose different solutions - the best of which were adopted.

Also, peer review is an integral part of open source development and open source mentality. If you prefer to be isolated at your ivory tower, well that's fine. I'm sure you can produce very high quality code. Just remember that even the best hunter misses a fox or two - I've seen some revered programmers (at different works I've been) making newbie mistakes - it's OK, their status quo is so high nobody can question their place, but even THEY benefit from letting their coworkers see their code and ask questions about it.