What a fantastic and paradoxical spin.
I agree, but I think all is not lost! Perhaps good estimations can be the result of incomplete theories that don't yet fully manage or describe all of the intrinsic complexity of the problem. We could lump complex ideas with simple, slightly larger analogs- like guessing the area of a crazy shape by drawing a square around it and measuring that rather than decomposing the interesting shape we're given.
Oh all is definitely not lost! As I've continued to think about this idea and related ideas I've come to believe that the key to estimation is to reduce the scope of what is being estimated. To shrink the "world" that the theory resides in. Not only does this have dramatic architectural consequences, but it also permits reasonably accurate estimates.
We're a place where coders share, stay up-to-date and grow their careers.
We strive for transparency and don't collect excess data.