Their match was played in November 2018. Here are the classical [1] and rapid [2] rating lists from that time. Fabiano was 12 points behind Magnus in classical, and 91 points behind him in rapid. It's all going to be relative. Obviously Fabiano is extremely good at all forms of chess, but all top players (in any sport) are going to have relative strengths/weaknesses when compared against each other.

[1] - https://web.archive.org/web/20181003104802/http://ratings.fi...

[2] - https://web.archive.org/web/20181009092541/https://ratings.f...

Classics is an extremely impractical format for ratings, because the games are so long so players can't play many games.

It's like Heisenberg Uncertainty.

Yet my impression is that the rapid list is volatile and often not reflecting the reality, the classical rating is better in tune with actual strength

It's the opposite - classical ratings is more reliable because there are regular organised events while there are very few serious blitz and rapid events. Most top players only play the rapid and blitz World Championship, no other events in these time controls.

You're right about the number of games influencing the rating volatility but the length of an individual game has nothing to do with it, it's about how many events are played how often.