The probability of a fruitfly's wing not being arrived at is 41000)-1/41000.
You think that says anything important? By far the majority of organisms do NOT have fruit fly wings. So?
So far, you've done nothing but, is effect, declare the endpoint of a drunkard's walk to be the target destination of a second drunkard's walk and using the improbability of that outcome to claim the first one was too improbable to occur.
If evolution is not 'trying' to to produce those wings, if there is not a target (and here's a hint: EVOLUTION HAS NO TARGET!!!!), there is no target to miss. So saying the likelihood of 'missing' is near unity has no meaning whatsoever.