If (nearly) every implementation of a good idea turns out badly, is it still a good idea ? It smells like the No True Scotsman fallacy.