"Evolution" is like looking at the outcome of a coin toss and saying, "That coin was predisposed to turn up heads because, look, it did."
The bird has wings because nature killed those previous mutations that did not. Fine. That's a paradigm, a lens through which the results can be viewed.
So can evolution in short lifespan creatures in the lab, and by observing species in the wild, particularly "ring" species that are all alive right now.
Only mutation can be measured.
Yes. And DNA mutations cause changes in offspring. Accumulated changes are "evolution".
Only mutation can be measured. "Evolution", i.e., natural selection, cannot.
Natural selection is one proposed component of evolution, but the two are not synonymous.
This is one of those oft repeated errors. It is factually challenged. Differently honest. To the far left of the bell curve as to reliability.
More specifically, differential reproductive success is probably among the most studied and measured phenomena in biology.