What if they said they were measuring how many of test subjects where human and then explicitly say the measure humans as having above 1 meter tail.

Now they can, without being misleading at all, claim there were no humans in the last us census.

Your definition of measurement must be representative of the category you claim to be measuring.

It is misleading to say there are no humans in the American census when your measurement for human is non-representive just like it is misleading to say that hardly any Americans are healthy when you're measurement fails to capture most healthy subjects.

I'd go with "inaccurate" then.