*
One problem with using natural selection as an explanation for a phenomenon is that the applicability of NS to any particular instance is a metaphysical assumption, not an empirical discovery.
That is, in trying to explain some biological phenomenon (such as why the giraffe has a long neck) then it is not a matter of discovering by research that natural selection was in fact the explanation for the specific phenomenon - this is impossible; rather it is first a matter of assuming that natural selection was the explanation, then doing science on that basis ...
(Roughly speaking, 'doing science' by making hypotheses about how 'this' could have happened, then planning observations and experiments to test these hypotheses of how it could have happened - by seeing whether observations and experiments are consistent with the hypotheses. Essentially, the process of science works by checking for coherence and consistency in a domain according to a specified set of causes.)
*
The assumption that natural selection was the cause of a specific phenomenon is something that is not, and cannot be, tested - because this assumption is outside science, comes before science, structures scientific investigations.
Paradoxically, it is the metaphysical - hence extra-scientific - nature of the concept of natural selection which enables it to serve as a kind of substitute for religion.
*
This is because natural selection is assumed to be the explanation for phenomena in general - therefore once this metaphysical assumption has been made, once natural selection has been accepted as universally valid, then all further experiences and investigations are structured by this assumption.
Therefore the 'truth' of natural selection is, apparently, reinforced by anything and everything which happens from that point onwards; natural selection can never fail to explain anything, because all valid explanations are required to conform to natural selection.
*
The problem is that those who accept natural selection as a universal explanation also typically do not acknowledge the fact that natural selection is a metaphysical assumption; instead they want (somehow) to say that natural selection is a product of science - while, at the same time, having the property of structuring science...
They typically want to assert that natural selection is not just an assumption with universal applicability, but that this assumption is necessary - that it is irrational to reject this assumption.
They want to argue that any rational and informed person is compelled by 'evidence' to accept the universal validity of natural selection.
*
So they are actually assuming the universal truth of natural selection, but falsely believe that they have instead been compelled by evidence to accept the universal truth of natural selection - 'because' everything they regard as valid evidence apparently fits-in with the theory of natural selection.
*
Hence natural selection functions, among those who regard it as inevitably universal, as their religion - that is, the bottom-line explanation of reality; while at the same time such people deny that natural selection is a religion - precisely because it is a metaphysical assumption outside science which nonetheless regards itself as an empirical discovery within science.
*
Natural selection is regarded as the epitome of truth and validity, precisely because of this error of classification.
For those who come to treat it as the fundamental reality, natural selection disguises its true nature as a structuring assumption and instead masquerades as a multiply-validated discovery.
Consequently, universal natural selection feels like an objectively factual yet also un-dis-proveable religion - the perfect religion! - necessarily correct and the master key to explaining everything!
*