How is it even possible to have a perfect race? I remember Martin pardoe telling me once it was not possible and that if he ever had one he would retire from orienteering...
Ask Gustav Bergman, he knows.
But to be a bit more serious, it depends on the resolution of your measurement. If you are going to count every second then of course it's impossible. However, most would agree that setting unattainable goals is unwise. When I say perfect in this case I mean that I spiked all control circles and didn't take any obviously wrong route choices. Or put in a different way I probably have a threshold of about 10 to 15 seconds before I count a mistake. That seems like a more reasonable, but still very difficult, goal.
True, it doesn't happen very often