Dumb...
NSW has a userbase of 31M. Its best-selling game for the year sold about 2M. Thats around 7% of its entire user base.
PS5 has an install base of 4.9M. 7% of that is ~400k. Why should you expect the PS5 to be selling at over 3-4x the rate that NSW sells games in relation to their respective install bases? Cause that is what it would have to do to make the top 10. Even if the PS5 best selling game sells as well as the NSW best-selling game percentage-wise in relation to their install base, that game on the pS5 will still not break 400K sales.
Install base (it would be more correct talking about the momentum of a console) helps achieving the full sales potential of a game however the sales potential is inherent of a game, it doesn't translate proportionally to the install base of a console.
So in 2017 when Switch just launched and install base was lower than 3DS and PS4, 4 out 10 of the best selling games in Japan for the year were for Switch.
Taking a more drastic example, in 2015 2 out 10 of the best selling games in Japan that year were for a failure like WiiU which always had a small install base.
People buy in droves what they really want to play, not just what happened to be on the console they own as if the game itself doesn't matter.
If a hit software happens to be on a low install base console then it will drive the console adoption and even if the momentum of a system is beyond repairing (like WiiU) it will anyway max out the reach within the small install base (high attach rate).
Top 10 best selling games at retail in Japan during 2017:
Top 10 best selling games at retail in Japan during 2015: