Aiming gold with stock as reserves will likely produce better results in the long term than aiming stocks with gold as reserves, simply due to gold's higher volatility.
I remember doing some backtesting using CEF (gold+silver, data goes back to 1992) and VTSMX as reserves and the results were a lot more favourable than with the two assets reversed.
I suspect therefore that for your 1992 onwards data that stocks were more productive than gold and AIM'ing gold with 'cash' (stock) accumulating (broadly progressive increased weighting) over time resulted in the 'better' (more rewarding) outcome.