OP here. I seriously thought about giving up on the gravity implementation many times, but not because the coding logic was difficult. I was more concerned that the player wouldn't find it compelling as I had early feedback it was way too confusing from player's perspective and I was too close to the game to notice it.

In the end, as it turned out, it wasn't the gravity mechanic that was too confusing, it was how haphazardly I initially implemented it. After months of playtesting and tweaking the difficulty curve in the early levels, I finally hit the sweet spot (which was measured based on feedback from play testers - mainly family, friends and a few colleagues).