The observation that dark colored moths predominate in areas where industrial pollution has stained the trees where they rest is correct. Kettlewell's error was in attributing this to a mutation. It turns out that until the pollution changed one of the environmental factors there had been a relative equilibrium in the frequency of alleles for dark and light color in the population. After pollution stained the trees, it was simply a change in expression frequency. A change, btw, which is totally in accordance with evolutionary theory. Kettlewell had the mechanism wrong, not the observation and not the application of this change as evidence for evolution. This effect occurs in many populations where environmental change causes a change in allelic frequency. The fact that the "white moth" population is rebounding since the soot levels have been reduced and there had been no change in other populations not effected by pollution, provides even more proof of evolution - not less. Here's the point: if the environmental pressure had been continued over evolutionary timescales and over a broad enough area, natural selection would have eliminated all white moths from the population (although the trait may have remained as a recessive).