Classical time control has to go, at least for now?

Let's face it: chess fans don't like draws. They complained about them in 1920s (Alekhine vs. Capablanca), in the 1980s (first Karpov vs. Kasparov match), and especially now, after Carlsen and Caruana drew all 12 classical games.

Still, Carlsen defeated Caruana in rapid quite handily, so there is no danger of "draw death" of chess in general. However, in the classical game, top grandmasters seem to have entered a zone where they have enough time to calculate most of anything a human opponent can throw at them (like an updated, modern version of Tigran Petrosian), but not enough time (and probably memory capacity) to calculate inhuman chess-engine variations such as "forced mate in 30+ moves" in the Carlsen-Caruana game 6. So, as Vasily Smyslov liked to say, they're mostly "making 40 good moves, the opponent makes 40 good moves too, and they agree to a draw."

The next generation of chess players may get enough computer training to start seeing some of the "inhuman" variations (or maybe not), but what should we do now?

Perhaps the only feasible solution may be to further shorten the "classical" time control, giving players less time to think everything through. This, of course, will somewhat lower the game quality, but at least it's not asking them to "purposely play bad", as Reti put it in his article about the Alekhine-Capablanca match.

On the other hand, who am I to propose changes of such magnitude? My playing strength was probably Candidate Master at its peak, which was more than 20 years ago. Perhaps an actual grandmaster will say that my reasoning is stupid and I should stick to history instead of proposing chess reforms