Via Ron Jeffries's blog, a ten-point critique of evidenciary claims made in Craig Larman's Agile and Iterative Development: A Manager's Guide": Misstating the Evidence for Agile and Iterative Development, Chapter 6, Evidence.
I haven't read Larman’s book, but the article’s author, Isaac Gouy, gives specific quotes out of Larman’s work, then does a handy job refuting those points based on evidence Larman himself cites.
It should be noted I am not jumping in on Gouy’s side. I’ve a huge distrust of short snippets of quotes taken from larger works. We see our lousy media doing a great job of handily, purposely distorting the gist of an argument via Dowdification. (Note that’s done on either side, not just via MoDo.)
Regardless of the Dowdification factor, Gouy’s article makes for interesting reading. Do be sure to read Jeffries’s commentary on Gouy’s background on comp.software.extreme-programming. Jeffries emphatic non-committal says something on its own.