That is very interesting. The apparent failure of the 'change one variable' approach is remiscent of what we frequently see in molecular evolution debates where ID/creationists fail to understand that there are a plethora of possible protein conformations that can perform particular functions and an even larger set of possible DNA sequences that can produce such functional conformations. For proteins you could easily subtitute 'intelligent life' and for DNA 'universal constants'.
Presumably it is normally accompanied by the assumption that the variables must be such that human life like our own could evolve, which is apparently Peg's position.
TTFN,
WK