Irrelevant. The argument isn't about the duration of sleep. We're interested in the two techniques' effectiveness in helping people fall asleep, not in keeping them asleep.

Irrelevant. We’re not interested in people who have no trouble falling asleep. The argument is a comparison between two groups of people who both already have trouble falling asleep.

It doesn’t matter if these people have ever tried sleeping pills. We only care about the relative efficacy of the two techniques, which we have no reason to assume depends on prior use.

This weakens the argument. The two groups of people aren't starting from the same level of difficulty falling asleep. Since we aren't comparing two equivalent groups, we can't conclude that behavior modification will be more effective across the board. Because people who use sleeping pills have had more difficulty falling asleep, it's possible that behavioral techniques would not be very effective at all for them.

It could be true that people who use behavioral techniques chose to do so because they want to avoid drugs, and also still true that behavioral techniques are more effective than drugs. This doesn't weaken the author's argument.