> There is no practically usable helium near the top of the atmosphere.
The context of this discussion and the fourth word in that sentence is important. Something existing isn't the same as something being practically available. That graph isn't wrong, percentage wise, but it's missing both density an cost per liter that makes it relevant to this discussion.
Agreed, but there's a reasonable bounds for that future context that sets what's "practical", at the time these reserves would run out if not replaced (this discussion), with that context partially dictated by todays practical alternatives (which also happen to be the exact same as the original source of the reserves!).
For an idea of the difficultly, compared to not bleeding helium into the atmosphere (as the petrol companies do now): the atmospheric pressure at those elevations is around 1/10,000,000,000,000 of that at sea level. To fill one party balloon, you would need to capture something like 5,000,000,000,000 party balloons of that atmosphere. Note: math might be a relatively negligible couple orders of magnitude off.
The context of this discussion and the fourth word in that sentence is important. Something existing isn't the same as something being practically available. That graph isn't wrong, percentage wise, but it's missing both density an cost per liter that makes it relevant to this discussion.