Mainly I was refering to finding the source of a problem more than fixing one you've already located. I never have problems fixing bugs; The solution just presents itself to me on a silver platter. Its the finding of some of those little buggers thats usually the hardest part. The code just writes itself. ... That and I do fine without sleep. Hell, I'm running on none right now. What can I say? Its the weekend and I had better things to do.
Bad idea, I've found that if I spend more than fifteen minutes on a specific problem it's time for a new approach. If you stubbornly keep working at it you'll fix the problem eventually, but you'll also probably miss a better solution and fail to avoid the many bugs that you may introduce with your sub-optimal solution and lack of sleep.