In between these opposite extremes -- your language clearly indicates that you think this level of what we might call "aggravated bias" (hiring not just someone in your network; but hiring them over a more qualified candidate, presumed to exist and be interested in your opportunity) is commonplace, or something close to it.

Such that in your mind, "including referrals in your hiring funnel" == aggravated bias (in the sense above), basically.

As to the bias stats you liked to: the findings interesting, to say the least -- if they can be validated. Unfortunately the link to the original Payscale "study" seems to be broken (if we can call it that -- since remember, this is the work of a private company, with products to push).