Biologists in general and geneticists in particular have been thinking and writing about this sort of thing for almost a hundred years. If there is anything that you can come up with that has not been thought about and probably experimented on already, up to 50 or more years ago then I'll burn my hat.
No "evolutionist" has ever seriously claimed that all or even most mutations are beneficial. Some have no effect and many are detrimental and some don't matter all that much.
There are 20 amino acids that are coded for by the RNA. There are four different nucleotides in RNA. It takes three of these codons to define an amino acid. A three-nucleotide sequence is called a codon. Since any codon may have each of the three nucleotide positions filled by any of the four codons, that means there are 4 x 4 x 4 combinations, that is, 64.
So 64 combinations code for 20 amino acids. Subtract 4 for codons that mean start and stop, and you have 60 doing the actual coding. Do you see that there are 40 codons that are not needed? Now, open a new tab on your computer and search for the "genetic code". You will find that in many cases changing the third nucleotide, or in four or so cases the first has no effect on the amino acid being coded. This is a "substitution mutation". Yes, this is still a mutation, and is called "a silent mutation" because it has no effect.
Now if you change the first or second nucleotide, you will mostly get a change in the amino acid. That might have some effect. However, all active proteins have structural regions and active regions, it does not matter whether they are enzymes, antibodies, cytokines, lymphokines, cell surface receptors, protein hormones etc.
What all these have in common is that it almost does not matter what the amino acid sequence of the structural region is, as long as it does not interfere with the active region. Even in the active regions, some "conservative substitutions" may be made, such as glycine for alanine or valine for leucine. These are only two of several such substitutions. which might be made without altering the activity of the enzyme, antibody etc very much. In these cases problems occur if the structural region interferes with the active region or if a non-conservative substitution happens in an active region.
This is particularly obvious in genes that code for basic structures such as body size, growth rates and bilateral symmetry. Many of these genes are highly conserved and while the structural regions may be highly varied in amino acid sequence, the active regions have remained functionally the same in organisms as varied as rice and humans.
Another example is the common use of insulin by diabetics. Until fairly recently, cow or pig insulin was used, and both of these differ from human insulin, yet they still worked. In about 1978 it was found how to make human insulin in microorganisms and most insulin is now genetically engineered human insulin from microorganisms. Yet "synthetic" insulin where one or two amino acids have been deliberately changed such as Lispro have some uses, such as quicker action or longer persistence in humans.
On top of all that, most proteins are coded for by RNA sequences that give a protein longer than needed. Bits at both ends may be clipped off before the protein is put to use. If the mutation is in the part clipped off, there is usually no effect.
Now another form of mutation is called a frameshift, which happens when a nucleotide is removed, this usually cannot be translated into a protein or if it can be, makes something unrelated to the original and by comparison is useless nonsense. Organisms that suffer badly from this often die before being born. In humans, this affects about 20 to 50% of conceptions, the embryo often dies before the woman is even aware she is pregnant. If you are born and survive infancy and childhood, congratulations, you are likely to be genetically good enough.
Evolution works on the few mutations that are either good enough, silent or a few beneficial ones.