This does not weaken the argument, We don’t know whether people normally drink milk while by itself or while eating food. If the former, then that makes no difference, because the participants weren’t eating. If the latter, then maybe even more people need lactase!

This weakens the argument by casting doubt on whether the study cited represents people’s normal dairy digestive needs. Maybe most people produce enough lactase for their usual dietary habits, even if a whole liter of milk at once would cause some problems.

This does not weaken the argument because we don’t know anything about the age of the study participants, nor how that corresponds to the general population. If the participants were younger than average, that might be a problem, but we have no idea.

This does not weaken the argument, because it doesn’t tell us anything about how representative the study was of the general population. Sure, lactose intolerant people should take lactase—but we still don’t know if that means 50 million people.

This does not weaken the argument, firstly because we don’t know what kind of dairy people usually consume, and secondly because if anything it indicates a greater need for lactase than the study does. Either way, this doesn’t undermine the “at least 50 million people” claim.