I asked whether that "meaningfully change[s] the picture"--i.e. the picture on the labor/capital income split. And I asked whether that practice is "happening at scale." I wasn't asking whether it was happening at all, but rather at sufficient scale that it would meaningfully skew the numbers on labor/capital income split.
Your data proved my point: loans against assets account for only 1-2% of unrealized gains, so even if you counted that as income, it would be a rounding error in the evaluation of the labor versus capital share of income.