A good point, and a serious problem. When I was in high school debate (Lincoln-Douglas), I hated the degree to which the competition was really about jargon and citation of overwhelming but irrelevant "evidence." I think the tipping point was when somebody claimed that teaching religion in public schools would lead to an environmental catastrophe (and even more, it was purely an argument from authority).

At one point, I ran a case that relied on no empirical evidence whatsoever (however abhorrent that may sound here): it was a quasi-Aristotelian argument that if you accept the value in the first premise--I believe it was "knowledge"--then the remainder followed. The whole case was perhaps three minutes long, half the allowed time, and formatted to make the series of premises and conclusions very obvious.

Best I could tell, there was only one weak link in the argument that was easily debatable. I correctly guessed that the people I was debating were more used to listing "evidence" than arguing logic, and most people had absolutely no idea how to handle even clearly stated premises and conclusions.

I was arguing against the position I actually hold, which is why there was still a flaw in the argument, but it won the majority of the debates nonetheless. Sad, more than anything.

This "best argument" idea disconsiders the danger of one argument against an army http://lesswrong.com/lw/ik/one_argument_against_an_army/