to have perfect selection in one generation, you had to pay the price of the elimination of the other 9,999 in that same generation (destroying your population) for the evolutionary scenario to work. - Hooby
This is just gibberish.
I'm tempted to leave it there but I won't.
Take the example mentioned in the OP. (Did you bother to read it?)
In birds who see UV light a point mutation has occurred at position 90 on the SWS opsin gene changing the first letter from "AGC" which produces the amino acid serine to "TGC" which produces cysteine. This single substitution slightly changes the physical shape of the protein molecule changing its sensitivity from violet light with wavelengths of 405 nm to ultraviolet at 360-370 nm.
The first bird to experience that mutation would have the advantage of being better able to spot caterpillars with UV markings. Let's say it was part of a local breeding population of 10,000 birds.
All the other 9,999 birds would continue to survive and breed successfully, but our UV-sensing bird would on average leave more descendants than the others. The difference would only be measurable over many generations but the percentage of birds in the population with this mutation would gradually tend towards 100%.
Natural selection preserves advantageous changes in the gene-pool - unless of course the first UV-sensitive bird got eaten by a sparrowhawk before it got time to breed. Such is life.
Measuring the rate of mutations and their effects on a population is a field of research with some of the most amazing mathematical minds in science. What are the chances somebody who has never read a book about the evidence for evolution has spotted something basic that they all missed?