They are BS under the first definition.

And I think we've gotten quiet far from the context their statement was made in.

The premise was that intelligence increases self control/delayed gratification. From this, we've extrapolated that havard students will have on average a higher self control, because they've had the the privilege of a much better education, which caused their IQ scores to be quiet high.

Personally, I'd agree that this extrapolation is flawed, because both factors are only correlated with no causal link.