I feel like you’re taking a very roundabout way to describe the concept of sample bias. So long as it’s a random sample, this is accounted for in the statistics. Yes, we should still get more data all the same.