I'm actually not making an accusation of overconfidence; just pointing out that using qualified language doesn't protect against it. I would prefer language that gives (or at least suggests) probability estimates or degrees of confidence, rather than phrases like "looks like" or "many suggest".

ID theorists are more likely than evolutionary biologists to use phrases like "looks like" or "many suggest" to defend their ideas, because those phrases hide the actual likelihood of ID. When I find myself thinking, "it could be that X," instead of "because of A and B, X is likely," I suspect myself of being overconfident, and I apply the same heuristic to statements from other people.