The limits of evolution

The Acts & Facts of September 2008 had some interesting quotes from Werner Arber
The following are quotes from the article without added comments. (The footnote numbers have been deleted from the body of the text but are cited at the end.)

The most recent replication is by Lenski et al, who evaluated the changes in over 30,000 generations of E. coli, concluding that millions of mutations and trillions of cells were needed to produce the estimated two to three mutations required to allow cells to bring citrate into the cell under oxic conditions. This corresponds with Michael Behe's deductions that if one mutation is required to confer some advantage to an organism, this event is likely; if two are required, the likelihood is far less; but if three or more are required, the probability rapidly grows exponentially worse, from very improbable to impossible. Evolution by mutations for this reason has very clear limits.
…
Regarding major evolutionary questions, such as the origin of the information required for natural selection, Arbor wrote in his Nobel Prize speech that the answers so far proposed are often trivial or avoid the major questions facing Darwinism. He gave the example of using meaningless phrases such as "evolutionary driving forces" to explain how life evolved. As Arbor wrote, the claim that "more intensive research is needed to understand the apparent complexity of nature" is actually an admission of ignorance about the origin of complexity in the living natural world.9
…
For his study of mutations, Arber selected bacteria because they have short generation times (20 minutes vs. 20 years for humans) and therefore reproduce enormous numbers of progeny in only a few days. They also do not have sophisticated genetic repair mechanisms as do eukaryotes, allowing far more mutations to be expressed in their offspring. One of Arber's studies evaluated 10,000 generations of E. coli under various conditions, finding that "tremendous diversity accumulated within each population." The phenotypic change was very rapid for the initial 2,000 generations, but far slower for the subsequent 8,000 generations, conforming to the research on viruses that found the rate of fitness gain "decelerated significantly over time," as did the rate of nucleotide substitution.11 Arber concluded that genetic variety has definite limits, a finding carefully documented by Behe.7

This evidence indicates that the changes he observed in bacteria resulted almost solely from transposition and other types of chromosomal rearrangement, not mutations as required by macroevolution. This study provides clear evidence that the putative evolution observed in microorganisms is primarily, if not totally, a result of built-in mechanisms designed to produce genetic, and thus phenotypic, variety.

References
7.Behe, Michael. 2007. The Edge of Evolution. New York: The Free Press.
8.Arber, W. 1996. Molecular Mechanisms Promoting and Limiting Genetic Variation. In Di Castri, F. and T. Younes (eds.), Biodiversity, Science and Development: Towards a New Partnership. Wallingford, Oxon (UK): CAB International.
9.Arber, W, D. Nathans, and H. O. Smith. 1992. 1978 Physiology or Medicine, Nobel Lectures: Physiology or Medicine 1971-1980, 469-492.
10.Arber, W. 1991. Elements in Microbial Evolution. Journal of Molecular Evolution. 33: 4-12.
11. 11.Papadopoulos, D. et al. 1999. Genomic Evolution During a 10,000-Generation Experiment with Bacteria. Proceedings of the National Academy of Science USA. 96: 3807-3812.