Depending on the bug, I could do those 47 amino acid changes in about 6 months, so 6 is a piece of cake. This would use modern molecular biology techniques. To use only "natural" mutational events, it might take a year or two.
The point is well taken that you cannot determine probability based on 6 mutations all occurring at one time in one bug. However, given that the average bacterial culture contains 1,000,000,000 bacteria per ml, a liter would have 1,000,000,000,000 bacteria. An average mutational rate of 1/1,000,000, means that there are 1,000,000 mutants in a liter of bacteria. Digest these numbers and you will see how easy it is to get to a nylon eating bug in a short period of time.
How do you conclusively show that a change is due to a "mutation event" rather than to an existing gene simply delivering different output functionality (perhaps based upon environmental changes or on differences from other genes)?
Wouldn't you have to account for the ratio of "favorable mutations" to all mutations? Any idea what that ratio is in bacteria?