It's not a straw man at all. And it is a red herring to call it a straw man!
Let's play with a few numbers. Or number concepts.
Take a fish. Any fish you like. Let's get it to become an amphibian. Any amphibian you like.
Get that leg and hip joint going. Mutations are required for strength and lengthening of bones. The joint requires tendons, cartilege, bursa, extra circulatory and nerve extensions and corresponding connections in the brain for movement of the joint. That's for EVERY joint.
Lungs need to form and be useful. Bone and digestive structures are revamped.
How many mutations for all of that?
Now, talking about mutations and genetics, we know now that it is not 'one gene per characteristic'. The genes act in coordination with a very complex timing system. All this must remain functioning while various mutations are integrated into the system. The fish must not be at a disadvantage for either survival or evolution at any time. And others must be willing to mate with it.
As you yourself said, the vast majority of mutations are either silent or harmful. Many of those that are harmful are lethal.
This leaves, actually and mathematically, a maximum of one potentially beneficial mutation per every thousand others.
That beneficial must STAY in the population, meaning it must be dominant.
Then another mutation must not only be there after a thousand others, but it also must not only stay in the population and be dominant, but it must work in coordination with the first mutation!
This has never, EVER been seen to happen even in the lowly E.coli! Scientists have been working with that little bacteria for over a hundred years now. That's over 2.5 MILLION linear generations (we won't count sisters and cousins and such!). Every mutagent in the book has been thrown at them, and a few that aren't in any book. Results? Millions of dead E.coli. A few bigger. One small population with a slight variation in the metabolic pathway. In 2.5 generations, none of them were ever identified as anything but E.coli.
That's 2.5 million generations of a prokaryote. No evolution no matter how hard it was attempted.
Let's talk time. There are several thousand, to say the least, mutations necessary to get that fish to become an amphibian. Each of those mutations must be considered in context of at least a thousand other mutations the population had to undergo and either marginalize or simply survive. Keep in mind that if there is one negative mutation per generation the entire population goes into error catastrophe and is therefore on an irreversible path to extinction.
Therefore we have to double the time the fish needs to adapt to circumstances which are forcing it to become an amphibian.
Most decent sized fish require at least a year generation time (egg to egg-laying maturation). Every 2000 years, then, you could count on a beneficial mutation. At least mathematically.
How many mutations are needed? Several thousand is a minimum and probably foolish minimum for so much more is needed. Three thousand beneficial mutations all working in concert with one another to produce an amphibian from a fish? Ridiculously small, but let's take a look.
3000
times 2000 years each
That's 6,000,000 years minimum to get from a fish to an amphibian.
Never mind how you got from proto-bacteria to a fish in the first place.
Never mind that reptiles, birds, and mammals all had to show up as well, involving millions of more mutations.
Never mind that the more complex you get, the longer the generation times get.
Ten to twelve years for the large primates; even longer for man.
Evolution has run out of time.
Especially if you consider that even the E.coli don't show ANY useful changes in 2.5 MILLION generations.
No, it's not a straw man. It's a valid and impossible argument for evolution.
Oh yeah, consider that it is the very lack of the millions of transitionals required that caused Gould and Eldridge to posit their views on fast, hidden evolution in small populations that left no trace in the fossil record.
Consider that no cell has ever been seen to make a de novo protein.
Consider that most proteins are frequently (daily on the average) disassembled by the cells and the amino acids recycled for use in other proteins being assembled.
Consider that new genetic information must make these new proteins and that the cell must also know what to do with them.
ALL of this above must happen, plus ever so much more, for evolution to proceed out of the bounds of simple variation within a kind.
Creation is the logical conclusion, since we have never seen any of the above!
It is evolution which requires blind faith in other evolutionists, a severe logic disability, ignoring of the evidence, and an inordinate amount of imagination and storytelling.
Cinderella and Sleeping Beauty have more truth in them!