Gentlemen, gentlemen, let us converse in harmony!
It is both rational to frame the sales in a more pessimistic manner to understand the lower bound of implications, and rational to frame sales in an optimistic manner to understand the upper bound of implications. Both are important for forecasting future possibilities, as nothing is truly inevitable in this world (barring death and taxes).
Anyway, VG Insights is better than Steam Spy and I checked their methodology and it’s better than Spy. Belgian Embassy’s number was very close to VGI. Right now, VGI has BG3 at 8m units, but has a standard deviation of 15% for 84% of its prior predictions, so, taking a pessimistic angle, 6.8M units sold is quite reasonable. Sales could be anywhere from 6.8 to 9.2. It is indeed no small feat for a CRPG, and a solid showing for a triple A game, but CRPGs tend to underperform, and BG3 is currently somewhat average by AAA standards.
The best way to frame the current sales numbers is not as a historical moment in video games, or a forgettable nothing burger, but as a decisive win for Larian Studios specifically.
