In what it was intended to convey, that cumulative selection for fitness-improving mutation makes a colossal difference to the rate at which high-fitness solutions evolve, yes I will continue to defend it.
Careful now. There are creationists who actually deny that natural selection even occurs at all. They don’t care about nor bother to consider concepts like fitness landscapes and their topology, they outright deny the very concept of fitness as being meaningful. That’s why you will find, for example, creationists deny that something like peppered moth colors have an effect on the rates of moth survival and reproductive success.
Here I believe I can confidently state that nobody actually thinks all evolutionary change occurs on perfectly smooth fitness landscapes. Rather, while it is true real fitness landscapes do have a more rugged feature to them, and that such features can reduce the rate at which adaptive evolution occurs to varying degrees, we know of no adaptation that sits in a prohibitively inhibitory position in the landscape, nor is there any indication that real fitness landscapes are so noisy and random that selection is rendered ineffective.
And we also know of evolutionary mechanisms that allow more efficient navigation on landscapes with more rugged features where epistatic effects of mutation affect the rate of adaptation.