The idea of avoiding obstacles by looking out of the window sounds exceptionally sensible to me, but I would be very wary of using 'alternative' performance data. The obstacle is only avoided vertically by 35' using normal data, so seeing it may not help very much!